Interactive collaborative navigation decision-making method and device, electronic equipment and storage medium

Abstract

The invention discloses an interactive collaborative navigation decision-making method and device, electronic equipment and a storage medium, and belongs to the technical field of marine navigation, and the method comprises the steps: determining a navigation decision-making scheme according to real-time target navigation data; generating a visual interpretation graph according to the decision scheme, the visual interpretation graph comprising a decision behavior graph generated according to the decision opportunity and decision behavior of the decision scheme, a navigation risk evolution graph generated according to the navigation risk of the decision scheme, and a feature contribution graph generated according to the correlation between the data features of the decision scheme and the decision scheme; a feature importance index graph is generated according to the feature contribution degree graph, and an inter-feature dependency degree graph is generated according to a non-linear relationship between the decision scheme features; and based on the visual interpretation graph, performing man-machine interaction cooperative judgment on the decision scheme, adjusting the decision scheme, and generating a final decision scheme. According to the method, the visual interpretation graph is generated according to the decision scheme, and the decision scheme and the visual interpretation graph are interactively and cooperatively judged and determined to determine the navigation decision.

Description

Technical Field

[0001] The present invention relates to the field of marine navigation technology, and in particular to an interactive collaborative navigation decision-making method, device, electronic equipment and storage medium. Background Art

[0002] With the continuous improvement of maritime navigation technology, how to remotely control the ship while ensuring navigation safety is the focus of modern technology research. Combining the real-time marine environment, the navigation status of the ship and the data of obstacles to quickly and accurately make navigation decisions is the key technology to ensure navigation safety.

[0003] In the existing technology, navigation behavior is determined by the historical driving experience of the pilot, which is overly dependent on the pilot's driving experience and cannot predict unknown risks. Different pilots may also have deviations in their situational cognition, making navigation decisions uncontrollable and posing a major safety hazard. There are also many intelligent machine models that can calculate the direction and speed of navigation through real-time monitoring data combined with various algorithms. However, in the machine learning process, due to the complexity of the maritime navigation environment, decision-making biases may occur, leading to navigation risks.

[0004] Therefore, there is an urgent need for an interactive collaborative navigation decision-making method that can provide a multi-faceted visual image interpretation of the decision scheme generated by the model so as to quickly and accurately optimize and adjust the decision scheme. Summary of the invention

[0005] In view of this, it is necessary to provide an interactive collaborative navigation decision-making method, device, electronic device and storage medium, which can perform multi-dimensional and multi-level visual image interpretation of the decision plan generated by the model to quickly determine the optimization adjustment of the decision plan.

[0006] In order to solve the above technical problems, on the one hand, the present invention provides an interactive collaborative navigation decision method, comprising: Based on TimesNet algorithm and deep learning network, the navigation decision plan is determined according to the real-time target navigation data; Generate a visual explanation diagram according to the decision scheme, the visual explanation diagram including a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the navigation risk of the decision scheme, a feature contribution diagram generated according to the data features of the decision scheme and the correlation of the decision scheme, a feature importance index diagram generated according to the contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; Based on the visual explanation diagram, the decision plan is judged collaboratively by human-computer interaction, the decision plan is adjusted, and a final decision plan is generated.

[0007] In one possible implementation, based on the TimesNet algorithm and deep learning network, a navigation decision plan is determined according to real-time target navigation data, including: Calculate the timing preference probability and behavior preference probability, timing deviation and behavior deviation, and navigation risk of historical navigation decision plans, and based on the preset decision evaluation function, select the decision plans that meet the preset thresholds to generate a training decision plan data set; Build an initial intelligent decision-making model based on TimesNet algorithm and deep learning network; Iteratively training the initial intelligent decision model according to the training decision solution data set to generate an intelligent decision model; Based on the intelligent decision model, a navigation decision plan is generated according to real-time target navigation data.

[0008] In a possible implementation, the timing preference probability and behavior preference probability, timing bias and behavior bias, and navigation risk of historical navigation decision plans are calculated, including: Extract the distribution characteristics of decision timing and decision behavior of historical decision plans, and set behavioral decision plan preferences based on the empirical distribution function; Calculate the timing preference probability and behavior preference probability of the historical decision plan according to the preset collision avoidance rules and the behavior decision plan preference; Determine the action range according to the preset collision avoidance rules, and determine the preset collision avoidance plan based on the relative side angle, the closest approach distance and the closest approach time; Calculating the decision timing deviation and decision behavior deviation between the preset collision avoidance scheme and the decision scheme in the decision library; According to the posterior Gaussian process and the probabilistic speed barrier method, the collision risk is determined, the navigation risk characteristic elements in the target navigation data are extracted, and the coupling relationship between the characteristic elements is calculated. The navigation risk is determined according to the coupling relationship between the collision risk and the characteristic elements.

[0009] In a possible implementation, based on a preset decision evaluation function, a decision solution whose decision evaluation function meets a preset threshold is screened out to generate a training decision solution data set, including: Based on the preset decision-making scheme indicators, the timing preference probability, behavior preference probability, timing deviation, behavior deviation and navigation risk index weights of the decision-making scheme are determined according to the CRITIC method; Construct decision evaluation function based on intuitionistic fuzzy sets and evidence theory; Calculate the decision evaluation value of the historical decision plan according to the decision evaluation function; Filter out decision plans whose decision evaluation values ​​meet the preset threshold and generate a training decision plan data set.

[0010] In a possible implementation, generating a visual explanation diagram according to the navigation decision scheme includes: Based on the intelligent decision model and SHAP algorithm, the feature contribution index, feature importance index and feature dependency index of the decision scheme are calculated according to the target navigation data; Using a visualization interface as a carrier, according to the interpretability of the feature contribution index, feature importance index and feature dependency index of the decision scheme, a feature contribution graph, a feature importance graph and a feature dependency graph are generated to form a feature interpretation graph; Using the visual interface as a carrier, a decision behavior diagram is constructed according to the decision timing and decision behavior of the decision plan; Using the visual interface as a carrier, a navigation risk evolution diagram is generated according to the navigation risk of the decision-making plan; The feature explanation diagram, decision behavior diagram and navigation risk evolution diagram constitute a visual explanation diagram.

[0011] In a possible implementation, based on the visual explanation diagram, the decision plan is judged by human-computer interaction, the decision plan is adjusted, and a final decision plan is generated, including: Based on the visual explanation diagram, judging the correctness of the decision plan according to the feature explanation diagram combined with the real-time monitoring data; When the decision timing and decision behavior of the decision plan deteriorate, the decision plan is adjusted in combination with the real-time monitoring data to generate a final decision plan; When the decision plan is consistent with the real-time monitoring data, no adjustment is made and the decision plan is the final decision plan.

[0012] In one possible implementation, it also includes: The modified final decision plan and its corresponding navigation data and environmental data are updated to the preset decision plan database; When the decision solutions in the decision solution database are updated to a preset threshold number, a training decision solution data set is screened out again according to the decision solutions in the decision solution database, and the intelligent decision model is iteratively updated.

[0013] In a second aspect, the present invention further provides an interactive collaborative navigation decision-making device, comprising: The decision-making scheme generation module determines the navigation decision scheme based on the real-time target navigation data based on the TimesNet algorithm and deep learning network; A visualization explanation diagram generation module generates a visualization explanation diagram according to the decision scheme, wherein the visualization explanation diagram includes a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the target navigation data, a feature contribution diagram generated according to the data features of the decision scheme and the correlation between the decision scheme, a feature importance index diagram generated according to the contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; The interactive collaboration module performs human-computer interactive collaborative judgment on the decision plan based on the visual explanation diagram, adjusts the decision plan, and generates a final decision plan.

[0014] In a third aspect, the present invention also provides an electronic device comprising a memory and a processor, wherein the memory is used to store programs and data; the processor is coupled to the memory, and is used to execute the programs stored in the memory to implement the interactive collaborative navigation decision method as described above, and / or to implement the interactive collaborative navigation decision as described above.

[0015] In a fourth aspect, the present invention further provides a computer storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the interactive collaborative navigation decision-making method as described above.

[0016] The beneficial effects of the present invention are as follows: first, based on the TimesNet algorithm and the deep learning network, the navigation decision plan is determined according to the real-time target navigation data; then, a visual explanation diagram is generated according to the decision plan, and the visual explanation diagram includes a decision behavior diagram generated according to the decision timing and decision behavior of the decision plan, a navigation risk evolution diagram generated according to the navigation risk of the decision plan, a feature contribution diagram generated according to the correlation between the data characteristics of the decision plan and the decision plan, a feature importance index diagram generated according to the feature contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision plan; finally, based on the visual explanation diagram, the decision plan is judged by human-computer interaction, the decision plan is adjusted, and the final decision plan is generated. The present invention performs multi-dimensional and multi-level visual graphic interpretation on the navigation decision plan generated by the deep learning network model, optimizes and adjusts the navigation decision plan according to these visual explanation diagrams, and obtains the final navigation decision plan. The present invention realizes fast and accurate real-time adjustment and optimization of the navigation plan from multiple aspects and levels to ensure navigation safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the 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 work.

[0018] Figure 1 A schematic diagram of a flow chart of an embodiment of the interactive collaborative navigation decision-making method provided by the present invention; Figure 2 For the present invention Figure 1 A schematic flow chart of an embodiment of step S101; Figure 3 For the present invention Figure 2 Schematic diagram of the process of the first embodiment of step S201; Figure 4 For the present invention Figure 2 FIG. 1 is a flow chart of a second embodiment of step S201; FIG. Figure 5 For the present invention Figure 1 A schematic diagram of a flow chart of an embodiment of step S102; Figure 6 For the present invention Figure 1 A schematic flow chart of an embodiment of step S103; Figure 7 A closed-loop schematic diagram of the interactive collaborative navigation decision-making process provided by the present invention; Figure 8 A schematic diagram of the structure of an embodiment of an interactive collaborative navigation decision-making electronic device provided by the present invention; Fig. 9 A schematic diagram of the structure of an embodiment of the interactive collaborative navigation decision storage medium provided by the present invention. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] In the description of the embodiments of the present invention, unless otherwise specified, "multiple" means two or more than two. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone.

[0021] The descriptions of "first", "second", etc. involved in the embodiments of the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the technical features defined as "first" or "second" may explicitly or implicitly include at least one of the features.

[0022] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0023] Before describing the embodiments, the following definitions are given for the relevant terms: The TimesNet algorithm is a time series analysis model based on deep learning, which aims to extract complex time changes by converting one-dimensional time series data into two-dimensional tensors and modeling them using two-dimensional convolution kernels. The SHAP algorithm is a method for explaining machine learning predictions by quantifying the impact of the output of a machine learning model on a specific sample. ‌‌

[0024] The present invention provides an interactive collaborative navigation decision-making method, device, electronic device and storage medium, which are described below respectively.

[0025] Figure 1 A flowchart of an embodiment of the interactive collaborative navigation decision-making method provided by the present invention is shown in FIG. Figure 1 As shown, the interactive collaborative navigation decision-making method includes: S101, based on TimesNet algorithm and deep learning network, determines the navigation decision plan according to real-time target navigation data; S102, generating a visual explanation diagram according to the decision scheme, the visual explanation diagram including a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the navigation risk of the decision scheme, a feature contribution diagram generated according to the data features of the decision scheme and the correlation of the decision scheme, a feature importance index diagram generated according to the feature contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; S103: Based on the visual explanation diagram, perform human-computer interactive collaborative judgment on the decision plan, adjust the decision plan, and generate a final decision plan.

[0026] It should be noted that an intelligent decision-making model is built based on the TimesNet algorithm and deep learning network, and the decision plans with excellent performance in historical navigation decision-making plans are screened out as training data sets for iterative training of the intelligent decision-making model. The trained intelligent decision-making model outputs navigation decision plans based on real-time navigation monitoring data including hull navigation data, marine navigation environment, and navigation data of target obstacles.

[0027] It should be further explained that the navigation decision plan must also comply with the "Regulations for Preventing Collisions on Inland Waters of the People's Republic of China". The navigation decision plan includes decision timing, decision behavior, navigation data and environmental data, as well as navigation risks predicted by the intelligent decision-making model.

[0028] This embodiment builds an intelligent decision model based on the TimesNet algorithm and the deep learning network, determines the navigation decision plan according to the intelligent decision model and the real-time target navigation data; generates a multi-level and multi-angle visual explanation diagram according to the decision plan, and the visual explanation diagram includes a decision behavior diagram generated according to the decision timing and decision behavior of the decision plan, a navigation risk evolution diagram generated according to the navigation risk of the decision plan, a feature contribution diagram generated according to the correlation between the data characteristics of the decision plan and the decision plan, a feature importance index diagram generated according to the feature contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision plan; based on these visual explanation diagrams, human-computer interaction and collaborative judgment are performed in combination with the decision plan, the decision plan is adjusted, and the final decision plan is generated. The present invention performs multi-dimensional and multi-level visual graphic interpretation on the navigation decision plan generated by the deep learning network model, optimizes and adjusts the navigation decision plan according to these visual explanation diagrams, and obtains the final navigation decision plan. The present invention realizes fast and accurate real-time adjustment and optimization of the navigation plan from multiple directions and levels to ensure navigation safety.

[0029] This embodiment performs multi-dimensional and multi-level visual graphic interpretation on the navigation decision plan generated by the deep learning network model, optimizes and adjusts the navigation decision plan based on these visual interpretation diagrams, and obtains the final navigation decision plan. The present invention realizes fast and accurate real-time adjustment and optimization of the navigation plan from multiple dimensions and levels to ensure navigation safety.

[0030] In some embodiments of the present invention, Figure 2 As shown, Figure 2 The present invention provides Figure 1 The flowchart of step S101 in an embodiment includes: S201, calculating the timing preference probability and behavior preference probability, timing deviation and behavior deviation, and navigation risk of historical navigation decision plans, and based on a preset decision evaluation function, screening out decision plans that meet a preset threshold, and generating a training decision plan data set; S202. Construct an initial intelligent decision-making model based on TimesNet algorithm and deep learning network; S203, iteratively training the initial intelligent decision model according to the training decision solution data set to generate an intelligent decision model; S204: Generate a navigation decision plan based on the intelligent decision model and the real-time target navigation data.

[0031] It should be noted that, based on the rich navigation experience of the control personnel, collision avoidance rules, navigation practice and navigation risks as evaluation indicators, the decision-making result evaluation criteria are established, the timing preference probability and behavior preference probability, timing deviation and behavior deviation and navigation risk of historical navigation decision-making plans are calculated, the weights of the above five indicators are determined based on the CRITIC method, and a decision evaluation function is constructed. According to the decision result evaluation criteria and the decision evaluation function, a decision result discriminator is constructed. According to the decision result discriminator, excellent navigation decision plans are screened out as training data sets. Among them, the larger the timing preference probability and behavior preference probability values, the more they are in line with navigation practice, the smaller the timing deviation and behavior deviation values, the more they are in line with collision avoidance rules, the smaller the navigation risk, and the better the decision result. These five indicators are combined with their indicator weights to screen out excellent decision plans.

[0032] This embodiment screens the navigation decision plans through five indicators, namely, timing preference probability and behavior preference probability, timing deviation and behavior deviation, and navigation risk, and selects a training decision plan data set with excellent performance, which allows the intelligent decision model to absorb excellent navigation experience and optimize the model, so that the navigation decision plan output by the intelligent decision model is more accurate.

[0033] In some embodiments of the present invention, Figure 3 As shown, Figure 3 The present invention provides Figure 2 The flowchart of the first embodiment of step S201 in the embodiment of calculating the timing preference probability and behavior preference probability, timing deviation and behavior deviation and navigation risk of the historical collision avoidance decision scheme includes: S301, extracting the distribution characteristics of decision timing and decision behavior of historical decision solutions, and setting behavioral decision solution preferences based on the empirical distribution function; S302, calculating the timing preference probability and behavior preference probability of the historical decision-making scheme according to the preset collision avoidance rules and the behavior decision-making scheme preference; S303, determining a behavior range according to a preset collision avoidance rule, and determining a preset collision avoidance plan based on a relative side angle, a closest approach distance, and a closest approach time; S304, calculating the decision timing deviation and decision behavior deviation between the preset collision avoidance solution and the decision solution in the decision database; S305. Determine the collision risk based on the posterior Gaussian process and the probabilistic speed barrier method, extract the navigation risk characteristic elements in the target navigation data, calculate the coupling relationship between the characteristic elements, and determine the navigation risk based on the collision risk and the coupling relationship between the characteristic elements.

[0034] It should be noted that the preset collision avoidance rules are in line with the "Rules for Preventing Collisions on Inland Rivers of the People's Republic of China" and are combined with navigation practices and navigation risks, with actual avoidance and avoidance behavior as rule indicators.

[0035] This embodiment calculates the timing preference probability and behavior preference probability, timing deviation and behavior deviation according to the decision timing and decision behavior of historical decision plans, combined with preset collision avoidance rules and the behavior decision plan preference, and calculates the navigation risk of historical decision plans, thereby providing an accurate and scientific basis for the subsequent screening of excellent historical decision plans.

[0036] In some embodiments of the present invention, Figure 4 As shown, Figure 4 The present invention provides Figure 1 The flowchart of the second embodiment of step S101 in the embodiment of the present invention is as follows: based on a preset decision evaluation function, a decision scheme whose decision evaluation function meets a preset threshold is screened out, and a training decision scheme data set is generated, including: S401. Based on the preset decision plan indicators, determine the timing preference probability and behavior preference probability, timing deviation and behavior deviation of the decision plan and the weight of the navigation risk indicator according to the CRITIC method; S402. Construct decision evaluation function based on intuitionistic fuzzy sets and evidence theory; S403, calculating the decision evaluation value of the historical decision plan according to the decision evaluation function; S403, screening out decision plans whose decision evaluation values ​​meet a preset threshold, and generating a training decision plan data set.

[0037] It should be noted that the larger the timing preference probability and behavior preference probability values ​​are, the more they conform to navigation practice; the smaller the timing deviation and behavior deviation values ​​are, the more they conform to the collision avoidance rules; the smaller the navigation risk is, the better the decision result is. These five indicators are combined with their indicator weights to screen out decision plans with excellent performance.

[0038] This embodiment constructs a decision evaluation function through the weights of timing preference probability and behavior preference probability, timing deviation and behavior deviation, and navigation risk indicators, and selects excellent decision plans as training data sets, so that the intelligent decision model absorbs excellent navigation experience and optimizes and improves the accuracy of the model.

[0039] In some embodiments of the present invention, Figure 5 As shown, Figure 5 The present invention provides Figure 1 The flowchart of an embodiment of step S102 in the embodiment includes: S501, based on the intelligent decision model and the SHAP algorithm, calculate the feature contribution index, feature importance index and feature dependency index of the decision solution according to the target navigation data; S502, using the visualization interface as a carrier, generating a feature contribution graph, a feature importance graph, and a feature dependency graph according to the interpretability of the feature contribution index, the feature importance index, and the feature dependency index of the decision solution to form a feature interpretation graph; S503, using the visualization interface as a carrier, constructing a decision behavior diagram according to the decision timing and decision behavior of the decision plan; S504, using the visual interface as a carrier, generating a navigation risk evolution diagram according to the navigation risk of the decision-making scheme; S505. The feature explanation diagram, decision behavior diagram and navigation risk evolution diagram form a visual explanation diagram.

[0040] It should be noted that in the navigation decision-making plan, each data is classified, and each feature represents a category, such as longitude and latitude, the distance between the hull and the target obstacle, sailing direction, sailing speed and wind speed, etc. At the same time, the influence mechanism between each feature is excavated through correlation analysis.

[0041] Specifically, the SHAP value of each feature in the decision-making scheme is calculated through the SHAP method in explainable AI, the importance index of each feature is determined, the dependency between features is calculated according to the influence mechanism between features, and the feature contribution index, feature importance index and feature dependency index are converted into visual images using the visualization interface as the carrier. At the same time, the decision-making timing and decision-making behavior of the decision scheme are converted into a decision-making behavior diagram, and the navigation risk of the decision scheme is converted into a navigation risk evolution diagram.

[0042] This embodiment uses the SHAP method in explainable AI to explain the characteristics of the decision-making plan and converts it into a visual image, thereby improving the transparency of the navigation decision and providing a multi-faceted and multi-level explanation for the decision-making plan.

[0043] In some embodiments of the present invention, Figure 6 As shown, Figure 6 The present invention provides Figure 1 The flowchart of an embodiment of step S103 in the embodiment includes: S601, judging the correctness of the decision plan based on the visual explanation diagram and the feature explanation diagram combined with the real-time monitoring data; S602, when the decision timing and decision behavior of the decision plan deteriorate, the decision plan is adjusted in combination with the real-time monitoring data to generate a final decision plan; S603. When the decision plan is consistent with the real-time monitoring data, no adjustment is made and the decision plan is the final decision plan.

[0044] It should be noted that the control personnel judge the decision-making plan based on the visual explanation diagram combined with the real-time monitoring data, determine whether the decision is timely, whether the decision-making behavior is correct, etc. The real-time monitoring data includes the navigation data of the ship, the marine environment data and the navigation data of the target obstacles. When the decision-making plan deviates from the actual monitoring data, the decision-making plan will be adjusted and optimized in time.

[0045] This embodiment uses visual explanation diagrams in combination with real-time monitoring data to adjust and optimize decision-making plans in a timely manner, thereby achieving human-computer interactive collaboration to make navigation decisions. At the same time, the provided visual explanation diagrams improve the navigation decision-making level of the control personnel, and the navigation plan is adjusted and optimized quickly and accurately in real time to ensure navigation safety.

[0046] In some embodiments of the present invention, the interactive collaborative navigation decision method further includes: The modified final decision plan and its corresponding navigation data and environmental data are updated to the preset decision plan database; When the decision solutions in the decision solution database are updated to a preset threshold number, a training decision solution data set is screened out again according to the decision solutions in the decision solution database, and the intelligent decision model is iteratively updated.

[0047] Specifically, the decision-making database stores historical final decision-making plans and historical human decision-making results. Each navigation decision-making plan will be updated in the database, and also includes the navigation data of the ship, the marine environment data, and the navigation data of the target obstacles corresponding to the decision-making plan. When the decision-making plan is updated to a certain number or a certain time, the training data set in the decision-making database is repeatedly filtered by the decision result discriminator according to actual needs, the model is iteratively updated, and the intelligent decision-making model is optimized.

[0048] Furthermore, excellent decision plans are screened out as model training data through the decision result discriminator, and good decision-making experience is fed back to the intelligent decision-making model through sample amplification to realize the iterative upgrade of the intelligent decision-making model, so that the intelligent decision-making model can output better decision plans, and the decision plan feature contribution graph, feature importance graph, feature dependency graph, decision behavior graph, navigation risk evolution graph, etc. output by the model are provided to the control personnel so that the control personnel can understand the navigation decision and adjust and optimize the decision plan, improve the level of navigation decision, and then realize human-computer interaction and collaboration, and update the final decision plan to the decision plan database, repeatedly iterate and update the model, and then form a closed loop, realizing the continuous iterative update of the model and the human-computer interaction and collaborative determination of navigation decisions, such as Figure 7 As shown, Figure 7 This is a closed-loop diagram of the interactive collaborative navigation decision-making process.

[0049] This embodiment improves the accuracy of the decision solutions output by the intelligent decision model by updating the final decision solution into the decision solution database and repeatedly iteratively updating the intelligent decision model.

[0050] In order to better implement the interactive collaborative navigation decision method in the embodiment of the present invention, based on the interactive collaborative navigation decision method, correspondingly, Figure 8 As shown, the embodiment of the present invention further provides an interactive collaborative navigation decision-making device 800 including: The decision-making scheme generating module 801 determines the navigation decision scheme according to the real-time target navigation data based on the TimesNet algorithm and the deep learning network; A visualization explanation diagram generation module 802 generates a visualization explanation diagram according to the decision scheme, wherein the visualization explanation diagram includes a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the target navigation data, a feature contribution diagram generated according to the data features of the decision scheme and the correlation between the decision scheme, a feature importance index diagram generated according to the feature contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; The interactive collaboration module 803 performs human-computer interactive collaborative judgment on the decision plan based on the visual explanation diagram, adjusts the decision plan, and generates a final decision plan.

[0051] The interactive collaborative navigation decision-making device 800 provided in the above embodiment can implement the technical solution described in the above interactive collaborative navigation decision-making method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above interactive collaborative navigation decision-making method embodiment, which will not be repeated here.

[0052] In the embodiment of the present invention, the interactive collaborative navigation decision-making device may be an independent server, or a server network or server cluster composed of servers. For example, the interactive collaborative navigation decision-making device described in the embodiment of the present invention includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets or a cloud server composed of a plurality of servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0053] The present invention also provides an interactive collaborative navigation decision-making electronic device, such as Fig. 9 As shown, Fig. 9The structure block diagram of an embodiment of the interactive collaborative navigation decision device provided by the present invention. The interactive collaborative navigation decision device 900 can be a computing device such as a mobile terminal, a desktop computer, a notebook, a palm computer, and a server. The interactive collaborative navigation decision device 900 includes a processor 901 and a memory 902, wherein the memory 902 stores an interactive collaborative navigation decision program 903.

[0054] In some embodiments, the memory 902 may be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory 902 may also be an external storage device of the computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device. Furthermore, the memory 902 may also include both an internal storage unit of the computer device and an external storage device. The memory 902 is used to store application software and various types of data installed in the computer device, such as program codes installed in the computer device. The memory 902 may also be used to temporarily store data that has been output or is to be output. In one embodiment, the interactive collaborative navigation decision program 903 may be executed by the processor 901, thereby realizing the interactive collaborative navigation decision method, device, electronic device and storage device of each embodiment of the present invention.

[0055] In some embodiments, the processor 901 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 902, such as executing interactive collaborative navigation decision-making programs.

[0056] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0057] The interactive collaborative navigation decision-making method, device, electronic device and storage device provided by the present invention are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. An interactive collaborative navigation decision-making method, characterized in that: include: Based on TimesNet algorithm and deep learning network, the navigation decision plan is determined according to the real-time target navigation data; Generate a visual explanation diagram according to the decision scheme, the visual explanation diagram including a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the navigation risk of the decision scheme, a feature contribution diagram generated according to the data features of the decision scheme and the correlation of the decision scheme, a feature importance index diagram generated according to the feature contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; Based on the visual explanation diagram, the decision plan is judged collaboratively by human-computer interaction, the decision plan is adjusted, and a final decision plan is generated.

2. The interactive collaborative navigation decision-making method according to claim 1, characterized in that: Based on TimesNet algorithm and deep learning network, the navigation decision plan is determined according to the real-time target navigation data, including: Calculate the timing preference probability and behavior preference probability, timing deviation and behavior deviation, and navigation risk of historical navigation decision plans, and based on the preset decision evaluation function, select the decision plans that meet the preset thresholds to generate a training decision plan data set; Build an initial intelligent decision-making model based on TimesNet algorithm and deep learning network; Iteratively training the initial intelligent decision model according to the training decision solution data set to generate an intelligent decision model; Based on the intelligent decision model, a navigation decision plan is generated according to real-time target navigation data.

3. The interactive collaborative navigation decision-making method according to claim 2, characterized in that: Calculate the timing preference probability and behavior preference probability, timing bias and behavior bias, and navigation risk of historical navigation decision plans, including: Extract the distribution characteristics of decision timing and decision behavior of historical decision plans, and set behavioral decision plan preferences based on the empirical distribution function; Calculate the timing preference probability and behavior preference probability of the historical decision plan according to the preset collision avoidance rules and the behavior decision plan preference; Determine the action range according to the preset collision avoidance rules, and determine the preset collision avoidance plan based on the relative side angle, the closest approach distance and the closest approach time; Calculating the decision timing deviation and decision behavior deviation between the preset collision avoidance scheme and the decision scheme in the decision library; According to the posterior Gaussian process and the probabilistic speed barrier method, the collision risk is determined, the navigation risk characteristic elements in the target navigation data are extracted, and the coupling relationship between the characteristic elements is calculated. The navigation risk is determined according to the coupling relationship between the collision risk and the characteristic elements.

4. The interactive collaborative navigation decision-making method according to claim 3, characterized in that: Based on the preset decision evaluation function, the decision solutions whose decision evaluation function meets the preset threshold are screened out, and a training decision solution data set is generated, including: Based on the preset decision-making scheme indicators, the timing preference probability, behavior preference probability, timing deviation, behavior deviation and navigation risk index weights of the decision-making scheme are determined according to the CRITIC method; Construct decision evaluation function based on intuitionistic fuzzy sets and evidence theory; Calculate the decision evaluation value of the historical decision plan according to the decision evaluation function; Filter out decision plans whose decision evaluation values ​​meet the preset threshold and generate a training decision plan data set.

5. The interactive collaborative navigation decision-making method according to claim 4, characterized in that: Generate a visual explanation diagram according to the navigation decision plan, including: Based on the intelligent decision model and SHAP algorithm, the feature contribution index, feature importance index and feature dependency index of the decision scheme are calculated according to the target navigation data; Using a visualization interface as a carrier, according to the interpretability of the feature contribution index, feature importance index and feature dependency index of the decision scheme, a feature contribution graph, a feature importance graph and a feature dependency graph are generated to form a feature interpretation graph; Using the visual interface as a carrier, a decision behavior diagram is constructed according to the decision timing and decision behavior of the decision plan; Using the visual interface as a carrier, a navigation risk evolution diagram is generated according to the navigation risk of the decision-making plan; The feature explanation diagram, decision behavior diagram and navigation risk evolution diagram constitute a visual explanation diagram.

6. The interactive collaborative navigation decision-making method according to claim 5, characterized in that: Based on the visual explanation diagram, the decision plan is judged by human-computer interaction, the decision plan is adjusted, and a final decision plan is generated, including: Based on the visual explanation diagram, judging the correctness of the decision plan according to the feature explanation diagram combined with the real-time monitoring data; When the decision timing and decision behavior of the decision plan deteriorate, the decision plan is adjusted in combination with the real-time monitoring data to generate a final decision plan; When the decision plan is consistent with the real-time monitoring data, no adjustment is made and the decision plan is the final decision plan.

7. The interactive collaborative navigation decision-making method according to claim 6, characterized in that: Also includes: The modified final decision plan and its corresponding navigation data and environmental data are updated to the preset decision plan database; When the decision solutions in the decision solution database are updated to a preset threshold number, a training decision solution data set is screened out again according to the decision solutions in the decision solution database, and the intelligent decision model is iteratively updated.

8. An interactive collaborative navigation decision-making device, characterized in that: include: The decision-making scheme generation module determines the navigation decision scheme based on the real-time target navigation data based on the TimesNet algorithm and deep learning network; A visualization explanation diagram generation module generates a visualization explanation diagram according to the decision scheme, wherein the visualization explanation diagram includes a decision behavior diagram generated according to the decision timing and decision behavior of the decision scheme, a navigation risk evolution diagram generated according to the target navigation data, a feature contribution diagram generated according to the data features of the decision scheme and the correlation between the decision scheme, a feature importance index diagram generated according to the contribution diagram, and a feature dependency diagram generated according to the nonlinear relationship between the features of the decision scheme; The interactive collaboration module performs human-computer interactive collaborative judgment on the decision plan based on the visual explanation diagram, adjusts the decision plan, and generates a final decision plan.

9. An interactive collaborative navigation decision-making electronic device, characterized in that: It comprises a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the steps of the interactive collaborative navigation decision-making method as claimed in any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores computer program instructions, which, when executed by a computer, enable the computer to execute the interactive collaborative navigation decision-making method according to any one of claims 1 to 7.

Images