A data processing method and device for marine search and rescue auxiliary decision-making

Abstract

This invention discloses a data processing method and apparatus for maritime search and rescue auxiliary decision-making. The method includes acquiring case information to be processed and a historical case dataset; preprocessing the case information to be processed and the historical case dataset to obtain preprocessed case information and a preprocessed historical case dataset; and processing the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set. As can be seen, this embodiment accurately identifies the historical search and rescue experience most similar to the current case through a multi-dimensional attribute similarity calculation method, obtaining the optimal matching case that comprehensively considers five attributes: time, space, numerical value, enumeration, and text. This enables rapid response and precise decision support for maritime emergencies, improving search and rescue efficiency, reducing risks, minimizing casualties and property losses, and providing intelligent and professional technical support for maritime search and rescue operations.

Description

Technical Field

[0001] This invention relates to the field of maritime search and rescue data processing technology, specifically to a data processing method and apparatus for auxiliary decision-making in maritime search and rescue. Background Technology

[0002] Case-based reasoning originated from Roger Schank's theory of "dynamic memory" (Schank, 1983). Its basic idea is to provide decision support for new problems by leveraging historical case solutions similar to the current one. This method retrieves the most similar historical cases from a case library and, combined with the specific characteristics of the current problem, directly reuses or modifies existing historical solutions, thereby achieving intelligent responses to the current problem.

[0003] Compared to the earlier start of research abroad, systematic research on case reasoning in China began in the late 1990s, with research directions mainly focusing on areas such as decision support, emergency early warning, fault diagnosis, and expert systems.

[0004] Case-based reasoning, as a mature and widely adaptable intelligent decision-making technology, has been effectively applied in various fields, including emergency management. However, case-based reasoning methods for maritime search and rescue still have significant shortcomings. These include the difficulty in accurately matching the multi-source heterogeneous features of historical cases in the maritime search and rescue field, the insufficient semantic understanding resulting from traditional case retrieval's reliance on keyword matching, and the lack of a dynamic adaptation mechanism for case attribute weight allocation. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a data processing method and apparatus for auxiliary decision-making in maritime search and rescue. By using a multi-dimensional attribute similarity calculation method, it accurately identifies the historical search and rescue experience most similar to the current case, and obtains the optimal matching case that comprehensively considers five types of attributes: time, space, numerical value, enumeration, and text. This enables rapid response and precise decision support for maritime emergencies, which is conducive to improving search and rescue efficiency, reducing risks, and reducing casualties and property losses, and provides intelligent and professional technical support for maritime search and rescue work.

[0006] To address the aforementioned technical problems, a first aspect of the present invention discloses a data processing method for auxiliary decision-making in maritime search and rescue, the method comprising:

[0007] S1, obtain information on cases to be processed and historical case datasets;

[0008] S2, preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information.

[0009] S3, process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

[0010] As an optional implementation, in a first aspect of the present invention, the preprocessing of the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset includes:

[0011] S21, Standardize the information of the case to be processed to obtain the first case information;

[0012] S22, perform missing value processing on the first case information to obtain the second case information;

[0013] S23, perform noise reduction processing on the second case information to obtain the third case information;

[0014] S24, Standardize the historical case dataset to obtain the first historical case dataset;

[0015] S25, perform missing value processing on the first historical case dataset to obtain the second historical case dataset;

[0016] S26, Denoise the second historical case dataset to obtain the third historical case dataset;

[0017] S27, perform unified format processing on the third case information and the third historical case dataset to obtain preprocessed case information and preprocessed historical case dataset.

[0018] As an optional implementation, in a first aspect of the present invention, processing the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set includes:

[0019] S31, the preprocessed case information and the preprocessed historical case dataset are processed to obtain case similarity information; the case similarity information includes first case similarity information, second case similarity information, third case similarity information, fourth case similarity information and fifth case similarity information;

[0020] S32, Analyze and process the preprocessed historical case dataset to obtain case success information and attribute weight information;

[0021] S33, process the case similarity information, the case success information, the attribute weight information and the preprocessed historical case dataset to obtain the target case information set.

[0022] As an optional implementation, in the first aspect of the present invention, the processing of the preprocessed case information and the preprocessed historical case dataset to obtain case similarity information includes:

[0023] S311, using the first case calculation model, the preprocessed time attribute information and the preprocessed historical case dataset are processed to obtain the first case similarity information; the first case similarity information includes date similarity information and time similarity information;

[0024] The calculation model for the first case is as follows:

[0025]

[0026] Wherein, RQ represents the date similarity information, SJ represents the time similarity information, and RQ i Let i be the date similarity values ​​in the date similarity information, representing the date similarity between the preprocessed case information and the i-th preprocessed historical case information in the preprocessed historical case dataset, SJ i For each of the i time similarity values ​​in the time similarity information, DIS represents the time similarity between the preprocessed case information and the i-th piece of preprocessed historical case information in the preprocessed historical case dataset. i The difference between the date in the preprocessing time attribute information and the date in the time attribute information of the i-th preprocessing historical case in the preprocessing historical case dataset, in days, is the DSS. i The absolute value of the difference between the time in the preprocessed time attribute information and the time in the time attribute information of the i-th preprocessed historical case in the preprocessed historical case dataset, in hours;

[0027] S312, The preprocessed spatial attribute information and the preprocessed historical case dataset are processed to obtain the second case similarity information;

[0028] S313, The preprocessed numerical attribute information and the preprocessed historical case dataset are processed to obtain the third case similarity information;

[0029] S314, The preprocessed enumerated attribute information and the preprocessed historical case dataset are processed to obtain the fourth case similarity information;

[0030] S315, the preprocessed textual attribute information and the preprocessed historical case dataset are processed to obtain the fifth case similarity information.

[0031] As an optional implementation, in the first aspect of the present invention, processing the preprocessed textual attribute information and the preprocessed historical case dataset to obtain fifth case similarity information includes:

[0032] S3151, Perform word vector conversion processing on the preprocessed text attribute information to obtain text word vectors;

[0033] S3152, The preprocessed historical case dataset is subjected to word vector transformation processing to obtain a historical text word vector set; the historical text word vector set includes several historical text word vectors;

[0034] S3153, Using the fourth case calculation model, the text word vectors and the historical text word vector set are calculated and processed to obtain text similarity information;

[0035] The calculation model for the fourth case is as follows:

[0036]

[0037] In the formula, WB represents the text similarity information. i YW represents the i-th text similarity value in the text similarity information, characterizing the similarity between the preprocessed case information and the text information of the i-th preprocessed historical case information in the preprocessed historical case dataset. LW represents the text word vector. i Let L be the i-th historical text word vector in the historical text word vector set, CosSim be the cosine similarity function, M be the dimension of the text word vector, and L be the length of the text word vector. i YW represents the dimension of the i-th historical text word vector in the historical text word vector set. j LW represents the value of the j-th dimension in the text word vector. i,k δ5 and δ6 represent the fifth and sixth weight parameters, respectively, in the i-th historical text word vector set.

[0038] S3154, The text similarity information is calculated and processed to obtain the fifth case similarity information. As an optional implementation, in the first aspect of this invention, the analysis and processing of the preprocessed historical case dataset to obtain case success information and attribute weight information includes:

[0039] S321, Perform a first analysis on the preprocessed historical case dataset to obtain first attribute weight information;

[0040] S322, Perform a second analysis on the preprocessed historical case dataset to obtain case success information and second attribute weight information;

[0041] S323, calculate and process the first attribute weight information and the second attribute weight information to obtain attribute weight information.

[0042] As an optional implementation, in a first aspect of the present invention, processing the case similarity information, the case success information, the attribute weight information, and the preprocessed historical case dataset to obtain a target case information set includes:

[0043] S331, The case similarity information and the attribute weight information are calculated and processed to obtain case matching information;

[0044] S332, Based on the case matching information, the preprocessed historical case dataset is filtered to obtain the filtered historical case dataset;

[0045] S333, Based on the success information of the cases, the filtered historical case dataset is processed to obtain the target case information set.

[0046] A second aspect of this invention discloses a data processing apparatus for auxiliary decision-making in maritime search and rescue, the apparatus comprising:

[0047] The acquisition module is used to acquire information about cases to be processed and historical case datasets;

[0048] The first calculation module is used to preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information.

[0049] The second calculation module is used to process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

[0050] A third aspect of this invention discloses another data processing apparatus for auxiliary decision-making in maritime search and rescue, the apparatus comprising:

[0051] processor;

[0052] A memory coupled to the processor stores executable program code;

[0053] The processor calls the executable program code stored in the memory to execute the data processing method for maritime search and rescue auxiliary decision-making disclosed in the first aspect of the present invention.

[0054] The fourth aspect of this invention discloses a computer-readable storage medium storing computer instructions, which, when invoked, are used to execute the data processing method for maritime search and rescue auxiliary decision-making disclosed in the first aspect of this invention.

[0055] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0056] In this embodiment of the invention, information on cases to be processed and a historical case dataset are acquired; the information on cases to be processed and the historical case dataset are preprocessed to obtain preprocessed case information and a preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumeration attribute information, and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information sets; the preprocessed case information and the preprocessed historical case dataset are processed to obtain a target case information set; the target case information set includes several target case information sets. It can be seen that this embodiment, through a multi-dimensional attribute similarity calculation method, accurately identifies the historical search and rescue experience most similar to the current case, obtaining the optimal matching case that comprehensively considers five types of attributes: time, space, numerical, enumeration, and text. This enables rapid response and precise decision support for maritime emergencies, which is beneficial for improving search and rescue efficiency, reducing risks, minimizing casualties and property losses, and providing intelligent and professional technical support for maritime search and rescue work. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a flowchart illustrating a data processing method for auxiliary decision-making in maritime search and rescue, as disclosed in an embodiment of the present invention.

[0059] Figure 2 This is a schematic diagram of the structure of a data processing device for maritime search and rescue auxiliary decision-making disclosed in an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram of another data processing device for maritime search and rescue auxiliary decision-making disclosed in an embodiment of the present invention. Detailed Implementation

[0061] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0063] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0064] This invention discloses a data processing method and apparatus for maritime search and rescue decision support. Through a multi-dimensional attribute similarity calculation method, it accurately identifies historical search and rescue experiences most similar to the current case, obtaining the optimal matching case that comprehensively considers five attributes: time, space, numerical value, enumeration, and text. This enables rapid response and precise decision support for maritime emergencies, improving search and rescue efficiency, reducing risks, minimizing casualties and property damage, and providing intelligent and professional technical support for maritime search and rescue operations. Detailed descriptions follow.

[0065] Example 1

[0066] Please see Figure 1 , Figure 1 This is a flowchart illustrating a data processing method for auxiliary decision-making in maritime search and rescue, as disclosed in an embodiment of the present invention. Figure 1 The described data processing method for maritime search and rescue auxiliary decision-making is applied in a data processing device for maritime search and rescue auxiliary decision-making, such as a local server or cloud server for data processing optimization management of maritime search and rescue auxiliary decision-making, etc., and the embodiments of the present invention are not limited thereto. Figure 1 As shown, the data processing method for maritime search and rescue auxiliary decision-making may include the following operations:

[0067] S1, obtain information on cases to be processed and historical case datasets;

[0068] It should be noted that the case information to be processed is the current maritime search and rescue site data information, which is obtained through sensors or various methods such as wireless and wired. Specifically, the embodiments of the present invention do not limit this. The case information to be processed includes time-related attributes: the date and time of the accident; spatial attributes: the geographical coordinates of the accident; numerical attributes: numerical information such as the number of people in distress; enumerated attributes: the type of accident (e.g., grounding, collision, sinking, etc.) and the accident level; and textual attributes: textual descriptions such as the type of vessel. Specifically, time-related attributes include the date and time of the accident; spatial attributes include the accident coordinates (represented by latitude and longitude); numerical attributes include the number of people in distress, wind speed, and wave height; enumerated attributes include the accident type, accident level, and wind direction; and textual attributes include the vessel type. This attribute-based classification method enables structured organization and efficient retrieval of data. Specialized similarity calculation methods are designed for different attribute characteristics, fully considering essential features such as time periodicity, spatial geographic characteristics, and textual semantic relationships. This facilitates hierarchical weight management and dynamic adjustment, supports multi-dimensional case representation and analysis, and constructs a unified processing framework, effectively improving the system's matching accuracy, adaptability, scalability, and maintainability, providing more comprehensive and reliable support for maritime search and rescue decision-making.

[0069] It should be noted that the case information to be processed has been structured using case representation methods before acquisition. Case representation refers to describing a case in a structured form that a computer can recognize and understand. The historical cases formed by this organization are the premise and foundation for case matching. Reasonable knowledge representation can effectively improve the efficiency and accuracy of case retrieval, and improve the quality of case adjustment and solutions. Based on the analysis of the characteristics of frame representation and object-oriented representation methods, this invention adopts a knowledge representation method that combines frames and object-oriented methods to describe maritime search and rescue cases, as shown in Table 1.

[0070] Table 1 Examples of knowledge representation in maritime search and rescue cases

[0071]

[0072]

[0073] It should be noted that the historical case dataset covers China's coastal waters, including the Bohai Sea, Yellow Sea, East China Sea, South China Sea, and the Pacific Ocean east of Taiwan Island, and contains information on all maritime accidents within my country's jurisdictional waters. The historical case dataset was sourced from the China Maritime Safety Administration website and collected using a focused web crawler, yielding 581 complete cases.

[0074] S2, preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information.

[0075] S3, process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

[0076] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0077] In an optional embodiment, the preprocessing of the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset includes:

[0078] S21, Standardize the information of the case to be processed to obtain the first case information;

[0079] It should be noted that the above standardization process can be carried out through methods such as max-min normalization, Z-score standardization, or fractional scaling standardization. In particular, the embodiments of the present invention do not limit the specific methods.

[0080] It should be noted that standardization can eliminate the influence of differences in the units of different attributes, making the values ​​of various attributes within a comparable range and improving the accuracy of subsequent similarity calculations.

[0081] S22, perform missing value processing on the first case information to obtain the second case information;

[0082] It should be noted that the above missing value processing can be carried out by means imputation, mode imputation or machine learning prediction imputation, etc., and the specific implementation of this invention is not limited.

[0083] It should be noted that by handling missing values, the problem of incomplete data in real-world maritime search and rescue scenarios can be solved, ensuring the integrity of case information and avoiding similarity calculation errors caused by missing data.

[0084] S23, perform noise reduction processing on the second case information to obtain the third case information;

[0085] It should be noted that the above-mentioned denoising process can be carried out by methods such as moving average, median filtering, wavelet transform, or outlier detection and replacement. In particular, the embodiments of the present invention do not limit the specific methods.

[0086] It should be noted that noise reduction processing can eliminate data noise introduced in marine environmental monitoring and manual data entry, improve data quality, and ensure the reliability of subsequent case matching.

[0087] S24, Standardize the historical case dataset to obtain the first historical case dataset;

[0088] It should be noted that the above standardization process can be carried out using the same standardization method as the information of the case to be processed, to ensure the consistency of data processing. Specifically, the embodiments of the present invention do not limit this.

[0089] It should be noted that standardization allows the historical case dataset and the case to be processed to be in the same numerical space, ensuring the fairness and comparability of similarity calculation.

[0090] S25, perform missing value processing on the first historical case dataset to obtain the second historical case dataset;

[0091] It should be noted that the above missing value handling can be carried out using the same missing value filling method as the information of the case to be processed, or a more suitable method can be selected according to the characteristics of historical data. Specifically, the embodiments of the present invention do not limit this.

[0092] It should be noted that handling missing values ​​can improve the completeness and representativeness of historical case datasets, thereby enhancing the knowledge value of the historical case library.

[0093] S26, Denoise the second historical case dataset to obtain the third historical case dataset;

[0094] It should be noted that the above denoising process can be carried out using the same denoising method as the information of the case to be processed, or a special data cleaning technique can be used to address the characteristics of historical data accumulation. Specifically, the embodiments of the present invention do not limit this.

[0095] It should be noted that noise reduction processing can improve the quality of historical case data, remove abnormal or erroneous data, and ensure the reference value of the case library.

[0096] S27, perform unified format processing on the third case information and the third historical case dataset to obtain preprocessed case information and preprocessed historical case dataset.

[0097] It should be noted that the above-mentioned unified format processing can be carried out through methods such as data structure transformation, attribute mapping or feature alignment to ensure that the current case and historical cases are consistent in data organization. Specifically, the embodiments of the present invention do not limit this.

[0098] It should be noted that by using a unified format, a standardized case representation model can be established, which facilitates subsequent batch similarity calculations and case matching, thereby improving the system's processing efficiency and scalability.

[0099] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0100] In an optional embodiment, the process of processing the preprocessed case information and the preprocessed historical case dataset to obtain the target case information set includes:

[0101] S31, the preprocessed case information and the preprocessed historical case dataset are processed to obtain case similarity information; the case similarity information includes first case similarity information, second case similarity information, third case similarity information, fourth case similarity information and fifth case similarity information;

[0102] It should be noted that the first case similarity information represents the degree of matching between the preprocessed case information and the preprocessed historical case dataset in terms of periodic temporal differences in the date and time of the accident; the second case similarity information represents the degree of spatial proximity between the preprocessed case information and the preprocessed historical case dataset in terms of the geographical coordinates of the accident; the third case similarity information represents the degree of proximity in the numerical differences between the preprocessed case information and the preprocessed historical case dataset in terms of numerical indicators such as the number of people in distress, wind speed, and wave height; the fourth case similarity information represents the degree of consistency in the categories between the preprocessed case information and the preprocessed historical case dataset in terms of classification indicators such as accident type, accident level, and wind direction; and the fifth case similarity information represents the degree of semantic relevance between the preprocessed case information and the preprocessed historical case dataset in terms of textual descriptions such as vessel type. By employing specialized similarity calculation methods for different attribute characteristics, the multidimensional matching relationship between the current case and historical cases can be comprehensively and accurately evaluated, providing reliable similar case support for the generation of maritime search and rescue plans.

[0103] S32, Analyze and process the preprocessed historical case dataset to obtain case success information and attribute weight information;

[0104] S33, process the case similarity information, the case success information, the attribute weight information and the preprocessed historical case dataset to obtain the target case information set.

[0105] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0106] In an optional embodiment, the process of processing the preprocessed case information and the preprocessed historical case dataset to obtain case similarity information includes:

[0107] S311, using the first case calculation model, the preprocessed time attribute information and the preprocessed historical case dataset are processed to obtain the first case similarity information; the first case similarity information includes date similarity information and time similarity information;

[0108] The calculation model for the first case is as follows:

[0109]

[0110] Wherein, RQ represents the date similarity information, SJ represents the time similarity information, and RQ iLet i be the date similarity values ​​in the date similarity information, representing the date similarity between the preprocessed case information and the i-th preprocessed historical case information in the preprocessed historical case dataset, SJ i For each of the i time similarity values ​​in the time similarity information, DIS represents the time similarity between the preprocessed case information and the i-th piece of preprocessed historical case information in the preprocessed historical case dataset. i The difference between the date in the preprocessing time attribute information and the date in the time attribute information of the i-th preprocessing historical case in the preprocessing historical case dataset, in days, is the DSS. i The absolute value of the difference between the time in the preprocessed time attribute information and the time in the time attribute information of the i-th preprocessed historical case in the preprocessed historical case dataset, in hours;

[0111] It should be noted that the first case calculation model cleverly solves the problem of the periodic characteristics of calendars and clocks by separating the date and time. i The formula takes into account the cyclical nature of dates throughout the year, achieving a smooth transition at the 182-day critical point; SJ i The formula specifically describes the similarity changes of time differences within a 24-hour cycle, accurately reflecting the actual proximity at different times, thus enabling precise matching of search and rescue opportunities. This model is computationally efficient and conforms to real-world application scenarios, accurately assessing the similarity of seasonal and diurnal environmental conditions during maritime search and rescue operations.

[0112] It should be noted that DIS i The DSS is obtained by calculating the difference in the number of days between two dates (ignoring the year, only calculating the difference between the month and day, ranging from 0 to 364 days). i The time periodicity is reasonably modeled by calculating the absolute value of the hour difference between two time points (after converting the time to 24-hour hours and subtracting the absolute value, the range is 0-23 hours).

[0113] S312, The preprocessed spatial attribute information and the preprocessed historical case dataset are processed to obtain the second case similarity information;

[0114] It should be noted that the above processing can be performed using geographical distance calculation methods such as Euclidean distance, Manhattan distance, and great circle distance, or it can be performed using the second case calculation model. Specifically, the embodiments of the present invention do not limit the specific processing.

[0115] The calculation model for the second case is as follows:

[0116]

[0117] In the formula, KJ represents the similarity information of the second case. i The similarity value of the i-th second case in the second case similarity information represents the spatiotemporal distance similarity between the preprocessed case information and the i-th preprocessed historical case information in the preprocessed historical case dataset. K and W are the longitude and latitude of the preprocessed spatial attribute information, respectively. i and E i These are the longitude and latitude of the spatial attribute information in the i-th preprocessed historical case information in the preprocessed historical case dataset, respectively; DVG represents the average spatiotemporal distance between the preprocessed case information and all the preprocessed historical case information in the preprocessed historical case dataset; R represents the Earth radius; and δ1, δ2, δ3 and δ4 represent the first weight parameter, the second weight parameter, the third weight parameter and the fourth weight parameter, respectively.

[0118] It should be noted that the first weight parameter, the second weight parameter, the third weight parameter, and the fourth weight parameter can be set by the user or obtained from historical data. Specifically, this embodiment of the invention does not limit the specifics.

[0119] It should be noted that the second case calculation model can accurately take into account the actual geographical distance under the influence of the Earth's curvature. At the same time, it achieves fine control over the distance calculation by introducing four adjustable weight parameters. Combined with the negative exponential normalization of distance, the similarity decreases non-linearly with the increase of distance. Furthermore, the use of average distance normalization avoids excessive penalties for long-distance cases, which is conducive to the accurate matching and evaluation of cases in different sea areas and at different distances in maritime search and rescue.

[0120] It should be noted that the first weight parameter (range 0.5-2) controls the overall influence of distance on similarity; the larger the value, the more significant the influence of distance on similarity. The second weight parameter (range 1-3) determines the rate at which similarity decays with increasing distance; the larger the value, the faster the similarity decreases with increasing distance. The third and fourth weight parameters (range 0.8-1.2) adjust the weights of latitude and longitude differences, respectively, allowing for regional adjustments based on the characteristics of different sea areas, thus improving the model's adaptability and accuracy in different sea environments.

[0121] S313, The preprocessed numerical attribute information and the preprocessed historical case dataset are processed to obtain the third case similarity information;

[0122] It should be noted that the above processing can be carried out by methods such as normalized Euclidean distance, Manhattan distance, Chebyshev distance or Mahalanobis distance, and the specific implementation of this invention is not limited.

[0123] It should be noted that by calculating the similarity of numerical attributes, the degree of difference in continuous numerical indicators such as the number of people in distress, wind speed, and wave height can be accurately assessed, providing a reliable basis for matching the scale of the distress with the marine environment and improving the targeting of search and rescue resource allocation.

[0124] S314, The preprocessed enumerated attribute information and the preprocessed historical case dataset are processed to obtain the fourth case similarity information;

[0125] It should be noted that the above processing can be carried out using methods such as Hamming distance, Jaccard coefficient, VDM (Value Difference Metric), or category similarity matrix based on domain knowledge. Specifically, the embodiments of the present invention do not limit the specific methods.

[0126] It should be noted that by calculating the similarity of enumerated attributes, the degree of matching of classification indicators such as accident type, accident level and wind direction can be accurately measured. This takes into account the inherent relationship between categories and provides precise guidance for the selection of search and rescue plans for specific types of accidents.

[0127] S315, the preprocessed textual attribute information and the preprocessed historical case dataset are processed to obtain the fifth case similarity information.

[0128] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0129] In an optional embodiment, the processing of the preprocessed textual attribute information and the preprocessed historical case dataset to obtain fifth case similarity information includes:

[0130] S3151, Perform word vector conversion processing on the preprocessed text attribute information to obtain text word vectors;

[0131] It should be noted that the above processing can be achieved through general word embedding methods such as Word2Vec, GloVe, and FastText, combined with pre-trained language models such as BERT and RoBERTa, as well as attention mechanisms or knowledge graph enhancement techniques. Alternatively, it can be achieved through a third-case computational model. In particular, the embodiments of this invention are not limited.

[0132] The calculation model for the third case is as follows:

[0133] WB=θ1·E(T)+θ2·D(T)+θ3·C(T);

[0134] θ1+θ2+θ3=1;

[0135] 0≤θ1,θ2,θ3≤1;

[0136] In the formula, WB is the text word vector, T is the preprocessed text attribute information, E(T) is the basic word embedding vector, D(T) is the domain adaptation vector, C(T) is the context enhancement vector, and θ1, θ2 and θ3 are the first weight factor, the second weight factor and the third weight factor, respectively.

[0137] It should be noted that the basic word embedding vector can be extracted through a pre-trained general word embedding model, which maps the ship type text to a general semantic space and captures basic semantic information; the domain adaptation vector can be obtained through a domain language model fine-tuned on the maritime search and rescue professional corpus, or through mapping and transformation of the domain dictionary and professional knowledge base, which enhances the professional semantic expression of ship type; the context enhancement vector can be calculated through an attention mechanism that integrates ship-related attribute information (such as purpose, structural characteristics, load capacity, etc.), or by semantic expansion using the associated information in the ship knowledge graph, which enriches the expression of the functional characteristics of ship type.

[0138] It should be noted that the first weighting factor controls the proportion of general semantic information, and can be appropriately increased when the ship type description is standardized and highly regulated; the second weighting factor adjusts the influence of specialized domain knowledge, and can be appropriately increased for ship type descriptions that are highly specialized and have obvious domain characteristics; the third weighting factor manages the integration of contextual information, and can be appropriately increased when the ship type description is simple but has rich related attribute information. The dynamic adjustment of the three weighting factors can be optimized according to the specific search and rescue scenario and data quality to achieve adaptive enhancement of word vector representation.

[0139] It should be noted that the third case calculation model can convert textual information such as ship type into a high-quality vector representation that combines general semantics, professional knowledge, and contextual characteristics. This breaks through the limitations of traditional word embedding methods that only focus on general semantics, fully considers the professional characteristics of the maritime search and rescue field and the functional attributes of ships, improves the domain adaptability and semantic richness of the text representation, provides a more accurate and comprehensive semantic foundation for subsequent ship type similarity calculation, and significantly enhances the accuracy of maritime search and rescue case matching.

[0140] S3152, The preprocessed historical case dataset is subjected to word vector transformation processing to obtain a historical text word vector set; the historical text word vector set includes several historical text word vectors;

[0141] It should be noted that the above processing can be carried out using general word embedding methods such as Word2Vec, GloVe, and FastText, combined with pre-trained language models such as BERT and RoBERTa, as well as attention mechanisms or knowledge graph enhancement techniques. Specifically, the embodiments of the present invention are not limited.

[0142] S3153, Using the fourth case calculation model, the text word vectors and the historical text word vector set are calculated and processed to obtain text similarity information;

[0143] The calculation model for the fourth case is as follows:

[0144]

[0145] In the formula, WB represents the text similarity information. i YW represents the i-th text similarity value in the text similarity information, characterizing the similarity between the preprocessed case information and the text information of the i-th preprocessed historical case information in the preprocessed historical case dataset. LW represents the text word vector. i Let L be the i-th historical text word vector in the historical text word vector set, CosSim be the cosine similarity function, M be the dimension of the text word vector, and L be the length of the text word vector. i YW represents the dimension of the i-th historical text word vector in the historical text word vector set. j LW represents the value of the j-th dimension in the text word vector. i,k δ5 and δ6 represent the fifth and sixth weight parameters, respectively, in the i-th historical text word vector set.

[0146] It should be noted that the fifth and sixth weight parameters can be set by the user or obtained from historical data. In particular, the embodiments of the present invention do not limit the specifics.

[0147] It should be noted that the fourth case calculation model uses cosine similarity to calculate the semantic similarity of ship type word vectors, and also introduces a vector quality assessment mechanism by detecting the proportions of zero-value elements XXS and YXS. i The assessment vector information completeness effectively addresses the sparsity problem of word vectors in text word vectors and historical text word vector sets. This model can identify and reduce the similarity weights of vectors with insufficient information, preventing erroneous matches caused by incomplete representations, and improving the robustness and adaptability of similarity calculation. In real search and rescue cases where there are non-standard and incomplete descriptions of ship types, it can significantly enhance the accuracy and reliability of case matching.

[0148] It should be noted that the fifth weight parameter ranges from [0.7, 1.4]. As an overall scaling factor, it controls the base weight of cosine similarity in the final similarity score. When δ5 > 1, the overall similarity value is increased, suitable for scenarios where the influence of text similarity needs to be enhanced; when δ5 < 1, the overall similarity value is decreased, suitable for scenarios where the weight of text factors needs to be reduced. Appropriately adjusting δ5 can balance the relative importance of text similarity in the overall case matching, improving the system's flexibility.

[0149] It should be noted that the sixth weight parameter ranges from [1, 8], and it controls the intensity of the penalty imposed by vector quality on similarity. When the value of δ6 is small, the penalty for zero elements in the vector is lighter, and the system has a high tolerance for incomplete information; when the value of δ6 is large, the requirements for vector quality are stricter, and the similarity of vectors with incomplete information will be significantly reduced. By adjusting δ6, the system can flexibly balance the accuracy and inclusiveness of similarity calculation according to the actual data quality and application requirements, effectively improving the model's adaptability under different data quality conditions.

[0150] S3154, The text similarity information is calculated and processed to obtain the fifth case similarity information.

[0151] It should be noted that the above processing normalizes text similarity information through methods such as min-max normalization and Z-Score normalization. The specific implementation of this invention is not limited.

[0152] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0153] In an optional embodiment, the step of analyzing the preprocessed historical case dataset to obtain case success information and attribute weight information includes:

[0154] S321, Perform a first analysis on the preprocessed historical case dataset to obtain first attribute weight information;

[0155] It should be noted that the above processing is carried out using methods such as the analytic hierarchy process, the Delphi method, expert scoring, principal component analysis, factor analysis, or fuzzy comprehensive evaluation. Specifically, the embodiments of the present invention are not limited to these methods.

[0156] It should be noted that by conducting expert-experience-oriented analysis on the preprocessed historical case dataset, the knowledge and experience of experts in the field of maritime search and rescue can be fully utilized to determine the relative importance of various attributes in case matching, form an attribute weighting system based on professional judgment, provide a reliable weighting basis for the comprehensive calculation of subsequent case similarity, and improve the professionalism and pertinence of search and rescue plan formulation.

[0157] S322, Perform a second analysis on the preprocessed historical case dataset to obtain case success information and second attribute weight information;

[0158] S323, Calculate and process the first attribute weight information and the second attribute weight information to obtain attribute weight information;

[0159] It should be noted that the above calculation process can be carried out by linear weighted average method or geometric weighted average method, or by the calculation model of the fifth case. Specifically, the embodiments of the present invention are not limited.

[0160] The calculation model for the fifth case is as follows:

[0161] QZ=ε1·QZY+(1-ε1)·QZE+ε2·(2·|ε1-0.5|-1)·CY;

[0162] -0.5≤ε1≤0.5;

[0163] In the formula, QZ represents the attribute weight information, QZY and QZE represent the first attribute weight information and the second attribute weight information, respectively, CY represents the difference information between the first attribute weight information and the second attribute weight information, and ε1 and ε2 represent the first fusion coefficient and the second fusion coefficient, respectively.

[0164] It should be noted that the second fusion coefficient can be set by the user or obtained from historical data. Specifically, this embodiment of the invention does not limit the specific value.

[0165] It should be noted that the fifth case calculation model, while considering the weighted average of the two weights in the fusion part, also introduces a difference enhancement term to capture key information at the point of weight divergence. This ensures that the attribute weights maintain the guidance of professional knowledge while also having the support of objective data, thereby significantly improving the accuracy and reliability of case matching and providing scientific and reasonable guidance on the importance of attributes for the generation of the final search and rescue plan.

[0166] It should be noted that the second fusion coefficient has a value range of [0,1]. It works together with the first fusion coefficient: the first fusion coefficient controls the relative proportion of the first attribute weight information and the second attribute weight information in the basic fusion. When the value is positive, it biases towards the first attribute weight information, and when it is negative, it biases towards the second attribute weight information; the second fusion coefficient adjusts the influence intensity of the difference term. The larger the value, the more attention is paid to the divergence between the two weighting methods. It also dynamically adjusts the directionality of the difference term according to the bias of the first fusion coefficient, realizing adaptive optimization of weight fusion and effectively balancing the contributions of subjective professional knowledge and objective data analysis in the determination of attribute weights.

[0167] It should be noted that CY can be obtained through algorithms such as the absolute difference vector method, the signed difference vector method, the Euclidean distance method, the relative difference vector method, and the variance-weighted difference method. Specifically, this embodiment of the invention does not limit the specific method used. When the signed difference vector method is used for calculation, CY = QZY - QZE, which can preserve the directional information of the differences, enabling the fusion process to distinguish between overestimation and underestimation of weights, thereby achieving more refined and targeted weight fusion adjustments.

[0168] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0169] In an optional embodiment, the second analysis processing of the preprocessed historical case dataset to obtain case success information and second attribute weight information includes:

[0170] S3221, Process all the preprocessed historical case information and case result information in the preprocessed historical case dataset to obtain case success information;

[0171] It should be noted that the above processing can be achieved through methods such as result scoring quantification, success rate statistics, effect classification, or multi-indicator comprehensive evaluation. Specifically, this embodiment of the invention does not limit the specific methods used. Through the above processing, the pre-processed historical case information and case results are quantitatively processed to evaluate multi-dimensional indicators such as personnel rescue rate, property preservation rate, and timeliness, thereby obtaining case success information. This case success information represents an objective evaluation score of the implementation effect of each historical search and rescue case, providing a standardized result measurement basis for evaluating the contribution of different attributes to search and rescue success.

[0172] S3222, Process the preprocessed historical case dataset and the case success information to obtain attribute correlation coefficient information;

[0173] It should be noted that the above processing can be obtained through statistical analysis methods such as Pearson correlation coefficient, Spearman rank correlation, mutual information, grey relational analysis, or information gain ratio. Specifically, this embodiment of the invention does not limit the specific methods used. Through the above processing, the statistical correlation between various attribute values ​​and successful case information is quantitatively calculated. These attribute values ​​include: time-based attribute values ​​(accident date, accident time), spatial attribute values ​​(accident coordinates), numerical attribute values ​​(number of people in distress, wind speed, wave height), enumerated attribute values ​​(accident type, accident level, wind direction), and textual attribute values ​​(ship type). The influence of changes in these different attributes on the search and rescue results is analyzed, thereby obtaining attribute correlation coefficient information. The attribute correlation coefficient information characterizes the influence strength of each of the 10 characteristic attributes on the success of the search and rescue case, quantifying the objective importance of each attribute in determining the success or failure of the search and rescue.

[0174] S3223, The attribute correlation coefficient information is normalized to obtain the second attribute weight information.

[0175] It should be noted that the above processing can be achieved through methods such as linear normalization, softmax function, coefficient of variation normalization, or range adjustment normalization. Specifically, the embodiments of the present invention do not limit the specific processing.

[0176] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0177] In an optional embodiment, the step of processing the case similarity information, the case success information, the attribute weight information, and the preprocessed historical case dataset to obtain the target case information set includes:

[0178] S331, The case similarity information and the attribute weight information are calculated and processed to obtain case matching information;

[0179] It should be noted that the above processing can be achieved through methods such as weighted summation, weighted geometric mean, weighted vector distance, or fuzzy comprehensive evaluation. Specifically, this embodiment of the invention does not limit the specific methods used. Through the above calculations, the five types of attribute similarity (temporal, spatial, numerical, enumeration, and textual similarity) are weighted according to their corresponding attribute weights to comprehensively evaluate the overall similarity between the current case and historical cases, thereby obtaining case matching information. This case matching information represents the comprehensive matching score between the preprocessed case information and each historical case in the preprocessed historical case dataset, providing a quantitative basis for case retrieval and selection.

[0180] S332, Based on the case matching information, the preprocessed historical case dataset is filtered to obtain the filtered historical case dataset;

[0181] It should be noted that the above processing can be achieved through methods such as similarity threshold filtering, Top-K selection, clustering grouping filtering, or adaptive similarity distribution filtering. Specifically, this embodiment of the invention does not limit the specific methods used. Through a case filtering algorithm, the preprocessed historical case information in the preprocessed historical case dataset is sorted and filtered according to case matching information. Several preprocessed historical case information with the highest matching degree are selected (exemplarily, 10 preprocessed historical case information), thus obtaining the filtered historical case dataset. The filtered historical case dataset represents the set of historical cases most similar to the preprocessed case information, narrowing the reference range for solutions and improving the targeting and efficiency of solution generation.

[0182] S333, Based on the success information of the cases, the filtered historical case dataset is processed to obtain the target case information set.

[0183] It should be noted that the above processing can be achieved through methods such as success rate-weighted selection, multi-objective optimization selection, or success rate threshold filtering. Specifically, this embodiment of the invention does not limit the specific methods used. Through the above processing, the cases in the filtered historical case dataset are comprehensively evaluated and ranked based on their historical success rates. Priority is given to selecting several preprocessed historical case information items that have both high similarity and high historical success rates (exemplarily, 5 preprocessed historical case information items), thereby obtaining the target case information set. The target case information set represents the reference cases and their solutions most suitable for the preprocessed case information.

[0184] As can be seen, the data processing method for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby enabling rapid response and precise decision support for maritime emergencies.

[0185] Example 2

[0186] Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a data processing device for auxiliary decision-making in maritime search and rescue, as disclosed in an embodiment of the present invention. Figure 2 The described data processing device for maritime search and rescue auxiliary decision-making is applied to a data processing optimization system for maritime search and rescue auxiliary decision-making, such as a local server or cloud server for data processing in maritime search and rescue auxiliary decision-making, etc., and the embodiments of the present invention are not limited thereto. Figure 2 As shown, the data processing device for maritime search and rescue auxiliary decision-making includes:

[0187] Module 201 is used to acquire information about cases to be processed and historical case datasets;

[0188] The first calculation module 202 is used to preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes a number of preprocessed historical case information.

[0189] The second calculation module 203 is used to process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

[0190] As can be seen, the data processing device for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby providing rapid response and precise decision support for maritime emergencies.

[0191] Example 3

[0192] Please see Figure 3 , Figure 3 This is a schematic diagram of another data processing device for maritime search and rescue auxiliary decision-making disclosed in an embodiment of the present invention. Figure 3 The described data processing device for maritime search and rescue auxiliary decision-making is applied to a data processing optimization system for maritime search and rescue auxiliary decision-making, such as a local server or cloud server for data processing in maritime search and rescue auxiliary decision-making, etc., and the embodiments of the present invention are not limited thereto. Figure 3 As shown, the data processing device for maritime search and rescue auxiliary decision-making includes:

[0193] Processor 301;

[0194] A memory 302 containing executable program code is coupled to the processor 301;

[0195] The processor 301 calls the executable program code stored in the memory 302 to execute some or all of the steps of the data processing method for maritime search and rescue auxiliary decision-making in Embodiment 1.

[0196] As can be seen, the data processing device for maritime search and rescue auxiliary decision-making described in the embodiments of the present invention can accurately identify historical search and rescue experiences most similar to the current case, thereby providing rapid response and precise decision support for maritime emergencies.

[0197] Example 4

[0198] This invention discloses a computer-readable storage medium storing computer instructions. When the computer instructions are invoked, they are used to execute some or all of the steps of the data processing method for maritime search and rescue auxiliary decision-making according to Embodiment 1.

[0199] Example 5

[0200] This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform some or all of the steps in the data processing method for maritime search and rescue auxiliary decision-making described in Embodiment 1.

[0201] The system embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0202] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0203] Finally, it should be noted that the data processing method and apparatus for maritime search and rescue auxiliary decision-making disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention, and are only used to illustrate the technical solutions of the present invention, not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data processing method for auxiliary decision-making in maritime search and rescue, characterized in that, The method includes: S1, obtain information on cases to be processed and historical case datasets; S2, preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information. S3, process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

2. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 1, characterized in that, The preprocessing of the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset includes: S21, Standardize the information of the case to be processed to obtain the first case information; S22, perform missing value processing on the first case information to obtain the second case information; S23, perform noise reduction processing on the second case information to obtain the third case information; S24, Standardize the historical case dataset to obtain the first historical case dataset; S25, perform missing value processing on the first historical case dataset to obtain the second historical case dataset; S26, Denoise the second historical case dataset to obtain the third historical case dataset; S27, perform unified format processing on the third case information and the third historical case dataset to obtain preprocessed case information and preprocessed historical case dataset.

3. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 1, characterized in that, The process of processing the preprocessed case information and the preprocessed historical case dataset to obtain the target case information set includes: S31, the preprocessed case information and the preprocessed historical case dataset are processed to obtain case similarity information; the case similarity information includes first case similarity information, second case similarity information, third case similarity information, fourth case similarity information and fifth case similarity information; S32, Analyze and process the preprocessed historical case dataset to obtain case success information and attribute weight information; S33, process the case similarity information, the case success information, the attribute weight information and the preprocessed historical case dataset to obtain the target case information set.

4. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 3, characterized in that, The process of processing the preprocessed case information and the preprocessed historical case dataset to obtain case similarity information includes: S311, using the first case calculation model, the preprocessed time attribute information and the preprocessed historical case dataset are processed to obtain the first case similarity information; the first case similarity information includes date similarity information and time similarity information; The calculation model for the first case is as follows: Wherein, RQ represents the date similarity information, SJ represents the time similarity information, and RQ i Let i be the date similarity values ​​in the date similarity information, representing the date similarity between the preprocessed case information and the i-th preprocessed historical case information in the preprocessed historical case dataset, SJ i For each of the i time similarity values ​​in the time similarity information, DIS represents the time similarity between the preprocessed case information and the i-th piece of preprocessed historical case information in the preprocessed historical case dataset. i The difference between the date in the preprocessing time attribute information and the date in the time attribute information of the i-th preprocessing historical case in the preprocessing historical case dataset, in days, is the DSS. i The absolute value of the difference between the time in the preprocessed time attribute information and the time in the time attribute information of the i-th preprocessed historical case in the preprocessed historical case dataset, in hours; S312, The preprocessed spatial attribute information and the preprocessed historical case dataset are processed to obtain the second case similarity information; S313, The preprocessed numerical attribute information and the preprocessed historical case dataset are processed to obtain the third case similarity information; S314, The preprocessed enumerated attribute information and the preprocessed historical case dataset are processed to obtain the fourth case similarity information; S315, the preprocessed textual attribute information and the preprocessed historical case dataset are processed to obtain the fifth case similarity information.

5. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 4, characterized in that, The process of processing the preprocessed textual attribute information and the preprocessed historical case dataset to obtain the fifth case similarity information includes: S3151, Perform word vector conversion processing on the preprocessed text attribute information to obtain text word vectors; S3152, The preprocessed historical case dataset is subjected to word vector transformation processing to obtain a historical text word vector set; the historical text word vector set includes several historical text word vectors; S3153, Using the fourth case calculation model, the text word vectors and the historical text word vector set are calculated and processed to obtain text similarity information; The calculation model for the fourth case is as follows: In the formula, WB represents the text similarity information. i YW represents the i-th text similarity value in the text similarity information, characterizing the similarity between the preprocessed case information and the text information of the i-th preprocessed historical case information in the preprocessed historical case dataset. LW represents the text word vector. i Let L be the i-th historical text word vector in the historical text word vector set, CosSim be the cosine similarity function, M be the dimension of the text word vector, and L be the length of the text word vector. i YW represents the dimension of the i-th historical text word vector in the historical text word vector set. j LW represents the value of the j-th dimension in the text word vector. i,k δ5 and δ6 represent the fifth and sixth weight parameters, respectively, in the i-th historical text word vector set. S3154, The text similarity information is calculated and processed to obtain the fifth case similarity information.

6. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 3, characterized in that, The analysis and processing of the preprocessed historical case dataset to obtain case success information and attribute weight information includes: S321, Perform a first analysis on the preprocessed historical case dataset to obtain first attribute weight information; S322, Perform a second analysis on the preprocessed historical case dataset to obtain case success information and second attribute weight information; S323, calculate and process the first attribute weight information and the second attribute weight information to obtain attribute weight information.

7. The data processing method for auxiliary decision-making in maritime search and rescue according to claim 3, characterized in that, The process of processing the case similarity information, the case success information, the attribute weight information, and the preprocessed historical case dataset to obtain the target case information set includes: S331, The case similarity information and the attribute weight information are calculated and processed to obtain case matching information; S332, Based on the case matching information, the preprocessed historical case dataset is filtered to obtain the filtered historical case dataset; S333, Based on the success information of the cases, the filtered historical case dataset is processed to obtain the target case information set.

8. A data processing device for auxiliary decision-making in maritime search and rescue, characterized in that, The device includes: The acquisition module is used to acquire information about cases to be processed and historical case datasets; The first calculation module is used to preprocess the case information to be processed and the historical case dataset to obtain preprocessed case information and preprocessed historical case dataset; the preprocessed case information includes preprocessed time attribute information, preprocessed spatial attribute information, preprocessed numerical attribute information, preprocessed enumerated attribute information and preprocessed text attribute information; the preprocessed historical case dataset includes several preprocessed historical case information. The second calculation module is used to process the preprocessed case information and the preprocessed historical case dataset to obtain a target case information set; the target case information set includes several target case information.

9. A data processing device for auxiliary decision-making in maritime search and rescue, characterized in that, The device includes: processor; A memory containing executable program code is coupled to the processor; The processor calls the executable program code stored in the memory to execute the data processing method for maritime search and rescue auxiliary decision-making as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when invoked, are used to execute the data processing method for maritime search and rescue auxiliary decision-making as described in any one of claims 1-7.

Images