A method and system for recommending virtual drill scenarios for underwater robot dam inspection

Abstract

This invention relates to the field of underwater dam inspection technology, and discloses a method and system for recommending virtual training scenarios for underwater robot dam inspection. This invention combines an assessment attribute set and a training scenario dataset to conduct virtual training skills assessments for underwater robots in multiple training scenarios. This clearly verifies the underwater robot's inspection capabilities in different scenarios, making the training more targeted, and providing a comprehensive understanding of the trainees' mastery of various skills in different scenarios. Furthermore, by using multiple skill assessment results to predict untested scenarios, it is possible to understand the underwater robot's performance in new scenarios in advance, providing a reference for developing training plans and enhancing the system's adaptability and foresight. Finally, by combining real-world case attribute datasets, existing assessment results, and prediction results to determine target recommended scenarios, it comprehensively considers actual conditions and future needs, making the recommended scenarios more accurate and practical.

Description

Technical Field

[0001] This invention relates to the field of underwater dam inspection technology, specifically to a method and system for recommending virtual drill scenarios for underwater robot dam inspection. Background Technology

[0002] With the development of underwater robot technology, underwater robots have been increasingly used in underwater dam inspection in recent years. However, due to the complexity and high cost of operating underwater robots, operators require extensive training before actual operations. Given the difficulty and high cost of actual underwater robot operations, operators are often trained through virtual drills. However, current traditional virtual drills primarily use a fixed-step scoring system, which cannot scientifically and comprehensively assess the trainees' skill mastery, decision-making, and adaptability. Furthermore, the availability of numerous drill scenarios makes it difficult for trainees to quickly select a suitable one. These factors reduce the efficiency of virtual drill training.

[0003] Existing technologies can only reflect whether trainees have completed all tasks in a certain training scenario. They do not conduct in-depth processing and analysis of the trainees' various skills. Trainees cannot intuitively understand their mastery of various skills. At the same time, the system cannot select and recommend suitable training scenarios for each trainee based on their skill mastery. Summary of the Invention

[0004] In view of this, the present invention provides a method and system for recommending virtual training scenarios for underwater robot dam inspection, in order to solve the problems in the existing technology that trainees cannot intuitively understand their mastery of various skills, and that the system cannot select and recommend suitable training scenarios for each trainee based on their skill mastery.

[0005] In a first aspect, the present invention provides a method for recommending virtual training scenarios for underwater robot dam inspection, for use in an underwater robot virtual training system; the method includes:

[0006] The process involves acquiring a dataset of real-world case attributes, a set of assessment attributes, and a dataset of training scenarios for multiple scenarios to be practiced. The real-world case attribute dataset and the training scenario dataset are stored in a structured format in a scenario attribute database and associated with the corresponding scenarios. Based on the assessment attribute set, the training scenario dataset is used to conduct virtual drills on underwater robot dam inspection skills in multiple scenarios to be practiced, resulting in multiple skill assessment results. Based on these multiple skill assessment results, the assessment results for underwater robot dam inspection skills in multiple initial, unpracticed scenarios are predicted, resulting in multiple skill assessment prediction results. Based on the real-world case attribute dataset, the multiple skill assessment results, and the multiple skill assessment prediction results, a target recommended scenario for underwater robot dam inspection virtual drills is determined.

[0007] The underwater robot dam inspection virtual drill scenario recommendation method provided by this invention first obtains a real-case attribute dataset, which accurately reflects the actual situation of underwater robot dam inspection. The assessment attribute set can evaluate from different capability dimensions, providing a standard and framework for subsequent comprehensive evaluation. Simultaneously, by structurally storing the drill scenario dataset and associating it with corresponding scenarios, different drill scenarios can be quickly and accurately invoked and managed, providing rich scenario resources for underwater robot dam inspection virtual drills, avoiding the time wasted on repeatedly constructing scenarios, and greatly improving drill efficiency. Secondly, by combining the assessment attribute set and the drill scenario dataset to conduct underwater robot dam inspection virtual drill skill assessment in multiple scenarios, the inspection capabilities of the underwater robot in different scenarios can be clearly verified, making the drills more targeted. This allows for a comprehensive understanding of the drill personnel's mastery of various skills in different scenarios. Furthermore, by using the results of multiple skill assessments to predict un-practiced scenarios, the performance of the underwater robot in new scenarios can be understood in advance, providing a reference for formulating drill plans and enhancing the system's adaptability and foresight. Finally, by combining real-world case attribute datasets, existing assessment results, and prediction results to determine the target recommendation scenario, we can comprehensively consider the actual situation and future needs, making the recommended scenario more accurate and practical.

[0008] In one optional implementation, based on the assessment attribute set, a virtual exercise skills assessment for underwater robot dam inspection is conducted in multiple exercise scenarios using a training scenario dataset, resulting in multiple skills assessment results, including:

[0009] The test matrix is ​​determined based on the assessment attribute set; the first assessment association matrix and the first score matrix of the scenario to be practiced are obtained. The first assessment association matrix is ​​used to associate the practice assessment items with the practice skill assessment attributes, and the first score matrix is ​​used to record the score of each assessment item; based on the assessment attribute set, the test matrix, the first assessment association matrix and the first score matrix, the underwater robot dam inspection virtual practice skill assessment is carried out in multiple scenarios to be practiced using the practice scenario dataset, and multiple skill assessment results are obtained.

[0010] The underwater robot dam inspection virtual drill scenario recommendation method provided by this invention covers a set of assessment attributes encompassing various key skill indicators, ensuring the completeness and systematic nature of the evaluation and avoiding the limitations of single-dimensional assessment. Furthermore, the first assessment correlation matrix precisely correlates the drill assessment items with the drill skill assessment attributes, ensuring that each assessment item clearly corresponds to a specific skill requirement. This helps assessors accurately judge the participants' performance in different skills, providing a basis for targeted improvement. Simultaneously, the first scoring matrix records the scores for each assessment item, achieving a quantitative assessment of the drill skills. Therefore, by determining the test matrix based on the assessment attribute set and combining it with the first assessment correlation matrix and the first scoring matrix, a comprehensive assessment of the underwater robot dam inspection virtual drill skills can be conducted from multiple dimensions. Furthermore, utilizing the drill scenario dataset for assessment in multiple drill scenarios enhances the adaptability and practicality of the technology. Different drill scenarios can simulate various real-world situations, testing the participants' ability to cope in different environments, ensuring that the assessment results are more representative and reliable.

[0011] In one alternative implementation, determining the test matrix based on the assessment attribute set includes:

[0012] Obtain the direct relationships between each assessment attribute in the assessment attribute set and determine the adjacency matrix. Assessment attributes are divided into primary attributes and secondary attributes. Obtain the comprehensive relationships between each assessment attribute in the assessment attribute set and determine the reachability matrix. Obtain the pass / fail status of each secondary attribute in the assessment attribute set and determine the cognitive structure. Based on the adjacency matrix, reachability matrix, and cognitive structure, determine the test matrix.

[0013] The underwater robot dam inspection virtual exercise scenario recommendation method provided by this invention can intuitively display the direct relationships between various assessment attributes in the assessment attribute set by determining an adjacency matrix. Furthermore, dividing the assessment attributes into primary and secondary attributes allows for a more hierarchical understanding of the direct relationships between attributes at different levels, facilitating targeted management and optimization. Further, determining the reachability matrix enables the acquisition of comprehensive relationships between various assessment attributes in the assessment attribute set, revealing not only direct connections but also indirect influence paths. Moreover, determining the cognitive structure allows for the acquisition of the pass / fail status of each secondary attribute in the assessment attribute set, thus providing a concrete understanding of whether each secondary attribute's performance in the current system meets the standards. Finally, combining the adjacency matrix, reachability matrix, and cognitive structure to determine the testing matrix ensures that the testing content is comprehensive, scientific, and targeted. Therefore, by comprehensively considering the information provided by the three matrices, the key relationships between various assessment attributes can be covered, while focusing on testing the pass / fail status of secondary attributes.

[0014] In one optional implementation, the assessment results of the underwater robot dam inspection virtual drill skills in multiple initial, unpracticed scenarios are predicted based on multiple skills assessment results, resulting in multiple skills assessment prediction results, including:

[0015] Obtain the second assessment correlation matrix and the second score matrix for multiple initial unpracticed scenarios; based on the multiple skill assessment results, the second assessment correlation matrix and the second score matrix, predict the assessment results of the underwater robot dam inspection virtual drill skill in multiple initial unpracticed scenarios, and obtain multiple skill assessment prediction results.

[0016] The underwater robot dam inspection virtual drill scenario recommendation method provided by this invention utilizes multiple existing skill assessment results for prediction, making full use of existing data and experience to discover potential patterns and commonalities between different scenarios. Simultaneously, by combining a second assessment correlation matrix and a second scoring matrix, the analysis of unpracticed scenarios can be further refined, improving the accuracy and reliability of predictions. Therefore, by implementing this invention, and by combining multiple skill assessment results, the second assessment correlation matrix, and the second scoring matrix to predict the assessment results of multiple initial unpracticed scenarios, the coverage of underwater robot dam inspection virtual drill skill assessment can be greatly expanded.

[0017] In one optional implementation, based on a real-world case attribute dataset, multiple skills assessment results, and multiple skills assessment prediction results, a target recommended scenario is determined for the virtual exercise of underwater robot dam inspection, including:

[0018] Based on multiple skill assessment prediction results, multiple target unpracticed scenarios are identified from multiple initial unpracticed scenarios; based on real case attribute datasets, multiple scenario importance values ​​and multiple scenario similarity values ​​are calculated for multiple target unpracticed scenarios; based on multiple skill assessment prediction results, multiple scenario importance values, and multiple scenario similarity values, target recommended scenarios for underwater robot dam inspection virtual drills are determined.

[0019] The underwater robot dam inspection virtual exercise scenario recommendation method provided by this invention can fully utilize the predicted results of multiple skill assessments to screen out multiple unpracticed target scenarios with potential value and demand for further analysis. Furthermore, by combining real-world case attribute datasets to calculate the importance values ​​of multiple scenarios, a quantitative evaluation of scenarios from a practical application perspective can be achieved. Simultaneously, by combining real-world case attribute datasets to calculate the similarity values ​​of multiple scenarios, similarities and correlations between different scenarios can be identified. Finally, by combining the predicted results of multiple skill assessments, the importance values ​​of multiple scenarios, and the similarity values ​​of multiple scenarios, the recommended target scenarios for underwater robot dam inspection virtual exercises are determined. This method comprehensively considers multiple factors, resulting in more scientific and reasonable recommendation results. Therefore, by implementing this invention, both the predicted assessment results and the importance and similarity in practical applications are considered, better meeting actual recommendation needs.

[0020] In one optional implementation, based on a real-world case attribute dataset, multiple scene importance values ​​and multiple scene similarity values ​​are calculated for multiple target unpracticed scenarios, including:

[0021] Calculate the frequency weights and similarity values ​​of multiple attributes in a real-world case attribute dataset; calculate the importance values ​​of multiple scenarios based on the frequency weights of multiple attributes; and calculate the similarity values ​​of multiple scenarios using the similarity values ​​of multiple attributes based on expert experience and the analytic hierarchy process.

[0022] The present invention provides a method for recommending virtual training scenarios for underwater robot dam inspection. This method calculates the frequency weights of multiple attributes in a real-world case dataset, quantifying the importance of each attribute in the actual case. Simultaneously, calculating multiple attribute similarity values ​​helps determine the degree of similarity between different scenarios at the attribute level. Furthermore, combining the frequency weights of multiple attributes to calculate the importance of multiple scenarios allows for a holistic assessment of the importance of each un-trained scenario for underwater robot dam inspection. Finally, when calculating the similarity values ​​of multiple scenarios using multiple attribute similarity values, expert experience provides an intuitive understanding and judgment of the scenarios, while the analytic hierarchy process (AHP) systematically handles the complex relationships between multiple attributes, ensuring the accuracy and reliability of the similarity values.

[0023] Secondly, the present invention provides an underwater robot virtual training system for performing the underwater robot dam inspection virtual training scenario recommendation method of the first aspect or any corresponding embodiment described above; the system includes: an underwater robot virtual training assessment module and a virtual training scenario recommendation module.

[0024] The underwater robot virtual training system provided by this invention, through the underwater robot virtual training assessment module and the virtual training scenario recommendation module, can comprehensively consider the actual situation and future needs, and thus more scientifically and rationally determine the target recommended scenarios for underwater robot dam inspection virtual training.

[0025] Thirdly, the present invention provides a computer device, including: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the underwater robot dam inspection virtual exercise scenario recommendation method described in the first aspect or any corresponding embodiment.

[0026] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the underwater robot dam inspection virtual exercise scenario recommendation method described in the first aspect or any corresponding embodiment.

[0027] Fifthly, the present invention provides a computer program product, including computer instructions, which are used to cause the computer to execute the underwater robot dam inspection virtual exercise scenario recommendation method described in the first aspect or any corresponding embodiment above. Attached Figure Description

[0028] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0029] Figure 1 This is a structural block diagram of an underwater robot virtual training system according to an embodiment of the present invention;

[0030] Figure 2 This is a flowchart illustrating a method for recommending virtual drill scenarios for underwater robot dam inspection according to an embodiment of the present invention.

[0031] Figure 3 This is a flowchart illustrating another method for recommending virtual drill scenarios for underwater robot dam inspection according to an embodiment of the present invention;

[0032] Figure 4 This is a flowchart illustrating another method for recommending virtual drill scenarios for underwater robot dam inspection according to an embodiment of the present invention;

[0033] Figure 5 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0035] Virtual training systems for underwater robots allow operators to conduct virtual drills anytime, anywhere, improving their skills. However, if the system lacks suitable assessment functions, operators cannot quickly assess their mastery of various skills, nor can they readily select suitable training scenarios from a wide range of options, thus limiting the effectiveness of virtual training.

[0036] Furthermore, most existing underwater robot virtual training systems use virtual engines, such as Unity3D and UE, to create training simulation scenarios with simple functions. The few systems with training evaluation functions only complete tasks step by step in a certain scenario and score them gradually, without a training scenario recommendation function.

[0037] According to an embodiment of the present invention, a method for recommending virtual drill scenarios for underwater robot dam inspection is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0038] This embodiment provides a method for recommending virtual drill scenarios for underwater robot dam inspection, used for, for example Figure 1 The underwater robot virtual training system 1 shown is shown. Figure 2 This is a flowchart of a virtual simulation exercise method for underwater robot dam inspection according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0039] Step S201: Obtain the real case attribute dataset, the assessment attribute dataset, and the training scenario dataset for multiple training scenarios.

[0040] Specifically, the real-world case attribute dataset represents a representative set of attribute data extracted from actual underwater robot dam inspection cases. It can include underwater robot performance data, dam feature data, environmental data, inspection task execution data, etc. It can be obtained in real time by installing various sensors on the underwater robot, or it can be obtained from historical databases.

[0041] The assessment attribute set represents a collection of attributes used to evaluate the underwater robot dam inspection virtual drill skills. In this embodiment, the underwater robot dam inspection virtual drill skills of the drill participants are assessed from four aspects: drill organization and scheduling ability, inspection task processing ability, emergency response ability, and comprehensive management ability. The assessment attributes are divided into primary attributes and secondary attributes. A drill participant is considered qualified for a primary attribute if all secondary attributes within that primary attribute are qualified. The assessment attribute settings in this embodiment are shown in Table 1 below.

[0042] Table 1. Assessment Attributes

[0043]

[0044] Real-world case attribute datasets and training scenario datasets are stored in a structured manner in the scenario attribute database and associated with their corresponding scenarios. Specifically, the real-world case attribute datasets and training scenario datasets are stored in a structured manner, divided into numerical, symbolic, and entity types. Numerical types include water level height and water flow velocity; symbolic types include weather conditions and water quality conditions; and entity types include defect structures, which can include defect type, defect severity, and defect location. Scenario feature attribute information is stored in the scenario attribute database and associated with its corresponding scenario.

[0045] Step S202: Based on the assessment attribute set, use the exercise scenario dataset to conduct virtual exercise skills assessment for underwater robot dam inspection in multiple exercise scenarios, and obtain multiple skills assessment results.

[0046] Specifically, the acquired training scenario dataset can be used to conduct virtual drills of underwater robot dam inspection in multiple training scenarios. The various assessment attributes in each training process can be evaluated by referring to the multiple assessment attributes and evaluation criteria contained in the assessment attribute set described in step S101, so as to obtain multiple corresponding skill assessment results reflecting the trainees' mastery of various training assessment skills.

[0047] Step S203: Based on multiple skill assessment results, predict the assessment results of the underwater robot dam inspection virtual drill skills in multiple initial unpracticed scenarios, and obtain multiple skill assessment prediction results.

[0048] Specifically, by combining the assessment results of multiple skills in the already practiced scenarios, the assessment results of the underwater robot dam inspection virtual exercise skills in multiple initial unpracticed scenarios can be predicted. This allows for advance understanding of the skills assessment situation in multiple initial unpracticed scenarios, enabling targeted preparation and training, thereby reducing unnecessary exercise time and resource waste and improving exercise efficiency.

[0049] For example, a predictive model between skills assessment results and scenario features can be established by selecting appropriate machine learning algorithms, such as regression analysis, decision trees, and neural networks. Existing assessment results are used as training data, scene features are input, and the corresponding skills assessment values ​​are output. Furthermore, features of initial, unpracticed scenarios can be input into the trained predictive model to obtain multiple skills assessment prediction results. The predictive model can be trained using training data, and its parameters can be continuously adjusted to minimize prediction errors. Methods such as cross-validation can be used to evaluate model performance and perform parameter adjustments and model optimization.

[0050] Step S204: Based on the real case attribute dataset, multiple skill assessment results, and multiple skill assessment prediction results, determine the target recommended scenario for the virtual exercise of underwater robot dam inspection.

[0051] Specifically, by analyzing real-world case attribute datasets, we can obtain the structural characteristics of different dam bodies, the problems and solutions encountered in actual inspections, and the impact of environmental factors on inspections. By analyzing multiple skills assessment results, we can understand the strengths and weaknesses of trainees in different training scenarios. By analyzing multiple skills assessment prediction results, we can understand the possible situations that may occur in non-training scenarios.

[0052] Therefore, by combining real-world case attribute datasets, multiple skills assessment results, and multiple skills assessment prediction results, more suitable and representative training scenarios for virtual drills of underwater robot dam inspection can be recommended to trainees or operators in a more comprehensive and accurate manner.

[0053] Furthermore, identifying target recommended scenarios can provide a clear direction for the planning of virtual drills for underwater robot dam inspections. In addition, the recommended scenarios can be used to reasonably arrange drill time and resource allocation, thereby improving the effectiveness and efficiency of the drills.

[0054] The underwater robot dam inspection virtual exercise scenario recommendation method provided in this embodiment first obtains a real-case attribute dataset, which can realistically reflect the actual situation in underwater robot dam inspection. The assessment attribute set can evaluate from different capability dimensions, providing a standard and framework for subsequent comprehensive evaluation. Simultaneously, by structurally storing the exercise scenario dataset and associating it with corresponding scenarios, different exercise scenarios can be quickly and accurately invoked and managed, providing rich scenario resources for underwater robot dam inspection virtual exercises, avoiding the time wasted on repeatedly building scenarios, and greatly improving exercise efficiency. Second, by combining the assessment attribute set and the exercise scenario dataset to conduct underwater robot dam inspection virtual exercise skill assessment in multiple scenarios to be exercised, the inspection capabilities of the underwater robot in different scenarios can be clearly tested, making the exercise more targeted, and thus providing a comprehensive understanding of the exercise personnel's mastery of various skills in different scenarios. Furthermore, by using the results of multiple skill assessments to predict un-exercised scenarios, the performance of the underwater robot in new scenarios can be understood in advance, providing a reference for formulating exercise plans and enhancing the system's adaptability and foresight. Finally, by combining real-world case attribute datasets, existing assessment results, and prediction results to determine the target recommendation scenario, we can comprehensively consider the actual situation and future needs, making the recommended scenario more accurate and practical.

[0055] This embodiment provides a method for recommending virtual drill scenarios for underwater robot dam inspection, used for, for example Figure 1 The underwater robot virtual training system 1 shown is shown. Figure 3 This is a flowchart of a virtual simulation exercise method for underwater robot dam inspection according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps:

[0056] Step S301: Obtain the real-world case attribute dataset, the assessment attribute set, and the training scenario dataset for multiple scenarios to be practiced. For details, please refer to [link to relevant documentation]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0057] Step S302: Based on the assessment attribute set, use the exercise scenario dataset to conduct virtual exercise skills assessment and evaluation of underwater robot dam inspection in multiple exercise scenarios, and obtain multiple skills assessment and evaluation results.

[0058] Specifically, step S302 includes:

[0059] Step S3021: Determine the test matrix based on the assessment attribute set.

[0060] Specifically, the test matrix can include an adjacency matrix, an reachability matrix, and a cognitive structure.

[0061] Among them, the adjacency matrix is ​​used to describe the direct relationships between each assessment attribute; the reachability matrix is ​​used to describe the comprehensive relationships between each assessment attribute; and the cognitive structure is used to record the trainees' mastery of the secondary attributes.

[0062] Therefore, the corresponding test matrix can be determined by combining the obtained set of assessment attributes.

[0063] In some optional implementations, step S3021 above includes:

[0064] Step a1: Obtain the direct relationships between each assessment attribute in the assessment attribute set and determine the adjacency matrix.

[0065] Step a2: Obtain the comprehensive relationships between all assessment attributes in the assessment attribute set and determine the reachability matrix.

[0066] Step a3: Obtain the pass / fail status of each secondary attribute in the assessment attribute set and determine the cognitive structure.

[0067] Step a4: Determine the test matrix based on the adjacency matrix, reachability matrix, and cognitive structure.

[0068] Specifically, the adjacency matrix can be determined by obtaining the direct relationships between the assessment attributes in the assessment attribute set. If there is a direct relationship between the assessment attributes, it is recorded as "1" in the corresponding position of the adjacency matrix; otherwise, it is recorded as "0".

[0069] The adjacency matrix is ​​typically a two-dimensional matrix with the number of rows and columns equal to the number of assessment attributes. For example, if there are five assessment attributes, the adjacency matrix would be a 5×5 matrix. If assessment attribute A is directly related to assessment attribute B, then the position at row A, column B and row B, column A of the matrix is ​​marked as "1".

[0070] Furthermore, the reachability matrix can be determined by obtaining the comprehensive relationships between each assessment attribute in the assessment attribute set. If there is a direct, indirect, or self-related relationship between each assessment attribute, the corresponding position in the adjacency matrix is ​​recorded as "1", otherwise it is recorded as "0".

[0071] Furthermore, the cognitive structure is used to record the trainees' mastery of the secondary attributes. The ideal response model can be used for recording. If the corresponding assessment attribute is qualified, it is recorded as "1", which means mastery; otherwise, it is recorded as "0", which means not mastery. This binary recording method is simple and clear, and is convenient for statistics and analysis.

[0072] Furthermore, the records of "1" and "0" provide a clear visual indication of which aspects the trainees have mastered and which aspects still require further improvement.

[0073] Finally, by combining the obtained adjacency matrix, reachability matrix, and cognitive structure, the corresponding test matrix can be determined.

[0074] Step S3022: Obtain the first assessment correlation matrix and the first score matrix of the scenario to be practiced.

[0075] The first assessment association matrix is ​​used to associate the exercise assessment items with the exercise skill assessment attributes; the first scoring matrix is ​​used to record the score of each assessment item.

[0076] The exercise and assessment items set in this embodiment are shown in Table 2 below:

[0077] Table 2. Exercise and Assessment Items

[0078]

[0079] In an optional embodiment, the first assessment correlation matrix is ​​shown in Table 3 below:

[0080] Table 3. First Assessment Correlation Matrix

[0081]

[0082] Step S3023: Based on the assessment attribute set, test matrix, first assessment correlation matrix and first score matrix, use the exercise scenario dataset to conduct virtual exercise skills assessment and evaluation of underwater robot dam inspection in multiple exercise scenarios, and obtain multiple skills assessment results.

[0083] Specifically, when conducting virtual drills and skills assessments for underwater robot dam inspection in multiple drill scenarios using a drill scenario dataset, the trainees' mastery of various drill skills can be obtained by combining the assessment attribute set, test matrix, first assessment correlation matrix, and first score matrix, thus yielding multiple skills assessment results.

[0084] [α 11 α 12 α 13 α 21 α 22 α 23 α 24 α 25 α 26 α 31 α 32 α 33 α 41 α 42 ].

[0085] In an optional embodiment, a training project includes a combination of several assessment items. For example, the assessment items for training project A include 1 equipment scheduling task, 2 hoisting operations, 5 fine inspection tasks, and 3 detailed inspection tasks. If the trainee successfully completes 1 equipment scheduling task, 1 hoisting operation, 4 fine inspection tasks, and 3 detailed inspection tasks in training project A, the percentage of their score is 82% of the total workload of the project assessment and the percentage of the number of tasks completed. The assessment skill mastery corresponding to the assessment correlation matrix is ​​[-,-,1,0.5,-,-,0.8,0.8,1,-,-,-,-,-], that is, the equipment scheduling ability scores 1 point, the robot hoisting ability scores 0.5 points, the defect identification ability scores 0.8 points, the defect detection ability scores 0.8 points, the defect and obstacle handling ability scores 1 point, and other scores are zero.

[0086] Step S303: Based on multiple skill assessment results, predict the assessment results of the underwater robot dam inspection virtual drill skills in multiple initial unpracticed scenarios, and obtain multiple skill assessment prediction results.

[0087] Specifically, step S303 includes:

[0088] Step S3031: Obtain the second assessment correlation matrix and the second score matrix for multiple initial unpracticed scenarios.

[0089] Specifically, the detailed descriptions of the second assessment correlation matrix and the second score matrix can be found in step S3022, where the descriptions of the first assessment correlation matrix and the first score matrix are provided. They will not be repeated here.

[0090] Step S3032: Based on multiple skill assessment results, the second assessment correlation matrix, and the second scoring matrix, predict the assessment results of the underwater robot dam inspection virtual drill skills in multiple initial unpracticed scenarios, and obtain multiple skill assessment prediction results.

[0091] Specifically, based on the trainees' mastery of various training and assessment skills, and combined with the assessment correlation matrix and score matrix of assessment scenarios that have not been practiced, the trainees' scores in virtual training scenarios that have not been practiced can be predicted, and the expected score percentage can be calculated to obtain the predicted results of multiple skills assessments.

[0092] For example, the assessment items for Exercise Project B include 1 engineering vehicle dispatching task, 1 operator dispatching task, 1 equipment dispatching task, 2 hoisting operations, 1 hoisting operation, 5 fine inspection tasks, and 6 detailed inspection tasks. Given the personnel's skill mastery matrix [-,-,1,0.5,-,-,0.8,0.8,1,-,-,-,-,-], it is predicted that the personnel can complete 1 equipment dispatching task, 0.5 hoisting operations (expected value), 4 fine inspection tasks, and 6 detailed inspection tasks in Exercise Project B. Therefore, their predicted percentage of score is 68%.

[0093] Step S304: Based on the real-world case attribute dataset, multiple skill assessment results, and multiple skill assessment prediction results, determine the recommended target scenario for the virtual drill of underwater robot dam inspection. For details, please refer to... Figure 1 Step S204 of the illustrated embodiment will not be described again here.

[0094] The underwater robot dam inspection virtual drill scenario recommendation method provided in this embodiment comprehensively evaluates the underwater robot dam inspection virtual drill skills from multiple dimensions by determining a test matrix based on an assessment attribute set and combining it with a first assessment correlation matrix and a first scoring matrix. Simultaneously, by utilizing a drill scenario dataset to conduct assessments in multiple scenarios to be drilled, the adaptability and practicality of the technology are enhanced. Different drill scenarios can simulate various real-world situations, testing participants' coping abilities under different environments, ensuring that the assessment results are more representative and reliable. Furthermore, by combining multiple skill assessment results, a second assessment correlation matrix, and a second scoring matrix to predict the assessment results of multiple initial un-drilled scenarios, the coverage of the underwater robot dam inspection virtual drill skills assessment can be greatly expanded. Finally, by combining a real-world case attribute dataset, existing assessment results, and prediction results to determine the target recommended scenario, the method comprehensively considers actual conditions and future needs, making the recommended scenarios more accurate and practical.

[0095] This embodiment provides a method for recommending virtual drill scenarios for underwater robot dam inspection, used for, for example Figure 1 The underwater robot virtual training system 1 shown is shown. Figure 4 This is a flowchart of a virtual simulation exercise method for underwater robot dam inspection according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps:

[0096] Step S401: Obtain the real-case attribute dataset, the assessment attribute set, and the training scenario dataset for multiple scenarios to be practiced. For details, please refer to [link to relevant documentation]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0097] Step S402: Based on the assessment attribute set, utilize the exercise scenario dataset to conduct virtual exercise skills assessment for underwater robot dam inspection in multiple exercise scenarios, obtaining multiple skills assessment results. For details, please refer to... Figure 3 Step S302 of the illustrated embodiment will not be described again here.

[0098] Step S403: Based on multiple skill assessment results, predict the assessment performance of the underwater robot dam inspection virtual drill skills in multiple initial, unpracticed scenarios, and obtain multiple skill assessment prediction results. For details, please refer to... Figure 3 Step S303 of the illustrated embodiment will not be described again here.

[0099] Step S404: Based on the real case attribute dataset, multiple skill assessment results, and multiple skill assessment prediction results, determine the target recommended scenario for the virtual exercise of underwater robot dam inspection.

[0100] Specifically, step S404 includes:

[0101] Step S4041: Based on the prediction results of multiple skills assessments, identify multiple target unpracticed scenarios from multiple initial unpracticed scenarios.

[0102] Specifically, multiple initial unpracticed scenarios can be sorted from low to high according to the predicted score ratio of the current trainees in the corresponding scenarios, i.e., the predicted skill assessment results, and the practice scenarios with the predicted score ratio between [50%, 70%] can be selected as the corresponding target unpracticed scenarios.

[0103] Step S4042: Based on the real case attribute dataset, calculate the importance values ​​and similarity values ​​of multiple target unpracticed scenarios.

[0104] Specifically, multiple scenario importance values ​​can comprehensively assess the importance of each unpracticed scenario for underwater robot dam inspection; multiple scenario similarity values ​​can help identify patterns and commonalities among similar scenarios.

[0105] In some optional implementations, step S4042 above includes:

[0106] Step b1: Calculate the frequency weights of multiple attributes and the similarity values ​​of multiple attributes in the real case attribute dataset.

[0107] Specifically, real-world case attribute values ​​are extracted from the real-world case attribute dataset. For example, for a specific real-world case, its dam type is determined to be a concrete dam, its dam height is a specific value, and its water flow velocity is within a certain range.

[0108] Furthermore, the frequency weight of each attribute value in real-world work examples is calculated; the higher the weight, the greater the importance of the attribute.

[0109] (1) Count the number of times each attribute value appears in all real operation cases. For example, count how many cases are concrete dams, how many cases have dam heights within a specific range, etc.

[0110] (2) Calculate the frequency of each attribute value, that is, the number of times the attribute value appears divided by the total number of real work cases.

[0111] (3) Normalize the frequency of each attribute value so that the sum of the frequencies of all attribute values ​​is 1.

[0112] (4) The normalized frequency is the frequency weight of the attribute value. The higher the weight, the more important the attribute value is in the real-world case.

[0113] Furthermore, the scene attribute data is divided into numeric, symbolic, and entity types, and the similarity of each attribute value is calculated separately. Entity type sub-attributes can be further divided into numeric and symbolic types. Numeric attributes can be viewed as two points; the greater the distance calculated, the lower the similarity, and vice versa. Symbolic attributes determine the similarity value by comparing whether the symbol values ​​are the same.

[0114] Step b2: Calculate the importance values ​​of multiple scenarios based on the frequency weight values ​​of multiple attributes.

[0115] Specifically, the presence or absence of each attribute value in the training scenario is counted, and the weighted sum of the scenario attribute values ​​is calculated: if an attribute value exists in the scenario, its frequency weight is added to the sum; if it does not exist, it is not added. For example, for the target untrained scenario A, if the concrete dam attribute value exists in the scenario, and the frequency weight of the concrete dam attribute value is 0.4, then 0.4 is added to the weighted sum of scenario A.

[0116] Furthermore, normalization is performed to calculate the importance value of each scenario: the weighted sum of all unpracticed scenarios is normalized so that the sum of the importance values ​​of all scenarios is 1. The normalized weighted sum is the importance value of each scenario. The higher the importance value, the greater the importance of that scenario in the underwater robot dam inspection.

[0117] In an optional instance, the real-world assignment cases are broken down into the assignments and their frequency as listed in the assessment item table. For example, if there are a total of n real-world assignment cases, and the fine-check work occurs a total of n1 times, then the weight of the fine-check work is... If the fine inspection work in exercise scenario C occurs n1' times, then its weighted value is calculated as I1 = k1 * n1'. The importance value is obtained by summing the weighted values ​​of all the work that occurs in exercise scenario C.

[0118] Step b3: Based on expert experience and the analytic hierarchy process, calculate the similarity values ​​of multiple scenes using multiple attribute similarity values.

[0119] Specifically, by combining expert scores and using the analytic hierarchy process (AHP), the weights of each attribute value in the similarity calculation are set, and the similarity of the exercise scenarios is calculated by weighted summation.

[0120] In an optional embodiment, the weight set K of each scenario factor and the attribute set S of the training scenario A are obtained by combining expert and other methods. A The attribute set S of exercise scenario B B The difference equation for a certain attribute of exercise scenarios A and B is shown in the following relationship:

[0121]

[0122] In the formula: S A [x] represents the value of a certain attribute in exercise scenario A; S B [x] represents the value of a certain attribute in exercise scenario B;

[0123] Furthermore, the similarity equation for training scenarios A and B is shown in the following formula:

[0124] f(n)=∑K*f(S A ,S B )

[0125] In the formula: f(n) represents the similarity value between the training scenarios A and B.

[0126] Step S4043: Based on multiple skill assessment prediction results, multiple scene importance values, and multiple scene similarity values, determine the target recommended scene for the virtual exercise of underwater robot dam inspection.

[0127] Specifically, by comprehensively considering multiple skill assessment prediction results, multiple scenario importance values, and multiple scenario similarity values, the target recommended scenario for the virtual exercise of underwater robot dam inspection is determined. This approach can comprehensively consider multiple factors, taking into account both the predicted assessment results and the importance and similarity in actual applications, thus making the target recommended scenario more scientific, reasonable, and representative.

[0128] The underwater robot dam inspection virtual exercise scenario recommendation method provided in this embodiment fully utilizes multiple skill assessment prediction results to screen out several target un-exercised scenarios with potential value and demand for further analysis. Furthermore, by combining real-world case attribute datasets to calculate the importance values ​​of multiple scenarios, a quantitative evaluation of scenarios from a practical application perspective can be achieved. Simultaneously, by combining real-world case attribute datasets to calculate the similarity values ​​of multiple scenarios, similarities and correlations between different scenarios can be identified. Finally, by combining multiple skill assessment prediction results, multiple scenario importance values, and multiple scenario similarity values, target recommended scenarios for underwater robot dam inspection virtual exercises are determined, comprehensively considering multiple factors to arrive at more scientific and reasonable recommendation results. Therefore, by implementing this invention, both the predicted assessment results and the importance and similarity in practical applications are considered, better meeting actual recommendation needs.

[0129] This embodiment provides an underwater robot virtual training system, such as Figure 1 As shown, the underwater robot virtual training system 1 includes: an underwater robot virtual training assessment module 11 and a virtual training scenario recommendation module 12.

[0130] Specifically, the underwater robot virtual exercise assessment module 11 is used to conduct underwater robot dam inspection virtual exercise skills assessment. The specific process can be referred to in steps S301 to S302 above, and will not be repeated here.

[0131] Furthermore, the virtual exercise scenario recommendation module 12 is used to determine the target recommended scenario for the virtual exercise of underwater robot dam inspection. The specific process can be referred to the description of steps S403 to S404 above, and will not be repeated here.

[0132] The underwater robot virtual training system provided in this embodiment, through its underwater robot virtual training assessment module and virtual training scenario recommendation module, can comprehensively consider actual conditions and future needs, thereby more scientifically and rationally determining the target recommended scenarios for underwater robot dam inspection virtual training. Therefore, by implementing this invention, the problem is solved: most existing underwater robot virtual training systems use virtual engines such as Unity3D and UE to create training simulation scenarios with simple functions within the engine; even the few systems with training assessment functions only complete tasks step-by-step within a certain scenario, gradually scoring them, and lack training scenario recommendation functionality.

[0133] In one example, a method for virtual drill assessment and drill scenario recommendation for underwater robot dam inspection is provided. This method is used to analyze and calculate the degree of mastery of various skills demonstrated by the drill participants during the drill, so that the drill participants can understand their strengths and weaknesses. At the same time, based on the degree of mastery of various skills, the method analyzes and recommends suitable drill scenarios to the drill participants.

[0134] The entire technical solution is illustrated below with examples:

[0135] The attribute data for each case and scenario is stored in a structured manner, categorized into numerical, symbolic, and entity types. Numerical data includes water level and flow velocity; symbolic data includes weather conditions and water quality; and entity data includes defect structures, such as defect type, severity, and location. Scene feature attribute information is stored in a scene attribute database and associated with the corresponding scene.

[0136] The underwater robot virtual exercise assessment module includes:

[0137] The assessment attributes are determined. The underwater robot virtual exercise assessment module evaluates the exercise personnel's skills from four aspects: exercise organization and scheduling ability, inspection task handling ability, emergency response ability, and comprehensive management ability. The assessment attributes are divided into primary attributes and secondary attributes. An exercise personnel passes the assessment for a primary attribute if all secondary attributes within that primary attribute are qualified. The assessment attribute settings in this embodiment are shown in Table 1.

[0138] Determine the testing matrix. The testing matrix includes an adjacency matrix, an accessibility matrix, and a cognitive structure. The adjacency matrix describes the direct relationships between each assessment attribute. If an attribute has a direct relationship, it is recorded as "1" in the corresponding position of the matrix; otherwise, it is recorded as "0". The accessibility matrix describes the overall relationships between each assessment attribute. If two attributes have a direct, indirect, or self-related relationship, the corresponding position of the matrix is ​​recorded as "1"; otherwise, it is recorded as "0". The cognitive structure records the trainees' mastery of the secondary attributes using an ideal response model. If the corresponding assessment attribute is passed, it is recorded as "1", indicating mastery; otherwise, it is recorded as "0", indicating non-mastery.

[0139] The underwater robot virtual exercise assessment module is designed with an assessment correlation matrix and a scoring matrix. The assessment correlation matrix (Q matrix) is used to associate the exercise assessment items with the exercise skill assessment attributes. The assessment items set in this embodiment are shown in Table 2.

[0140] The scoring matrix is ​​used to record the scores for each assessment item. The underwater robot virtual exercise assessment module obtains information on the trainees' mastery of various exercise assessment skills through specific assessment attributes, test matrices, scoring matrices, and assessment correlation matrices.

[0141] The percentage of scores in the practice assessment items indicates the level of skill mastery.

[0142] For example, a training exercise may consist of a combination of assessment items. Exercise A may include one equipment scheduling task, two hoisting operations, five fine inspections, and three detailed inspections. If a participant successfully completes one equipment scheduling task, one hoisting operation, four fine inspections, and three detailed inspections in Exercise A, their score percentage (the percentage of the total workload and the number of tasks completed) is 82%. The resulting skill mastery correlation matrix is ​​[-,-,1,0.5,-,-,0.8,0.8,1,-,-,-,-,-], meaning: 1 point for equipment scheduling ability, 0.5 points for robot hoisting ability, 0.8 points for defect identification ability, 0.8 points for defect detection ability, 1 point for defect and obstacle handling ability, and no points for other skills.

[0143] The virtual drill scenario recommendation module includes:

[0144] Based on the assessment of the trainees' mastery of various assessment skills obtained from the underwater robot virtual exercise assessment module, and combined with the assessment correlation matrix and score matrix of the assessment scenarios that were not exercised, the trainees' scores in the virtual exercise scenarios that were not exercised are predicted, and the expected score percentage is calculated.

[0145] If the assessment items for Exercise Item B include 1 engineering vehicle dispatching task, 1 operator dispatching task, 1 equipment dispatching task, 2 hoisting operations, 1 hoisting operation, 5 fine inspection tasks, and 6 detailed inspection tasks, and given the personnel's skill mastery matrix [-,-,1,0.5,-,-,0.8,0.8,1,-,-,-,-,-], and predicting that the personnel can complete 1 equipment dispatching task, 0.5 hoisting operations (expected value), 4 fine inspection tasks, and 6 detailed inspection tasks in Exercise Item B, then their predicted percentage of score is 68%.

[0146] The exercise scenarios are sorted from low to high according to the predicted score ratio of the current exercisers in the corresponding scenarios, and exercise scenarios with predicted score ratios between [50%, 70%] are selected first.

[0147] The part of calculating the importance of the exercise scenarios first extracts the attribute values ​​of real operation cases, calculates the frequency weight of each attribute value in the real operation cases, and the higher the weight, the higher the importance of the attribute. Then, the presence or absence of each attribute value in the exercise scenarios is counted. Finally, the weighted sum of the scenario attribute values ​​is calculated and normalized to calculate the importance value of each scenario.

[0148] The real-world assignment cases are broken down into the tasks and their frequency in the assessment item table. For example, if there are a total of n real-world assignment cases, and the detailed inspection task occurs n1 times, then the weight of the detailed inspection task is... If the fine inspection work in exercise scenario C occurs n1' times, then its weighted value is calculated as I1 = k1 * n1'. The importance value is obtained by summing the weighted values ​​of all the work that occurs in exercise scenario C.

[0149] The section on calculating the similarity of training scenarios categorizes scenario attribute data into numeric, symbolic, and entity types, and calculates the similarity of each attribute value. Entity-type sub-attributes can be further divided into numeric and symbolic types. Numerical attributes can be viewed as two points; the greater the distance between them, the lower the similarity, and vice versa. Symbolic attributes determine similarity by comparing whether the symbol values ​​are the same.

[0150] Combining expert scores, the Analytic Hierarchy Process (AHP) is used to assign weights to each attribute value in the similarity calculation, and a weighted sum is applied to calculate the similarity of the training scenarios. Finally, by incorporating information from multiple factors, suitable training scenarios are recommended to the participants.

[0151] Furthermore, existing technologies can only improve overall training levels through repeated practice within scenarios, but they cannot identify individual strengths and weaknesses from the training results, nor can they provide targeted practice for skills that are not well mastered. At the same time, different training scenarios have different focuses, making it difficult for trainees to quickly select the scenario with the best training effect from a scenario library. The virtual training assessment and scenario recommendation method for underwater robot dam inspection provided in this example can analyze the trainees' current mastery of various skills based on their training results.

[0152] In an optional instance, the performance of the above-mentioned underwater robot dam inspection virtual exercise assessment and exercise scenario recommendation method is evaluated using marginal precision rate (MMR) and pattern precision rate (PMR) on underwater robot dam inspection virtual exercise.

[0153] Specifically, the marginal precision rate is used to calculate the average accuracy of individual attribute judgments, and the calculation formula is shown in the following relationship:

[0154]

[0155] In the formula: K represents the total number of attributes; N represents the number of participants in the test; n k This indicates the number of people whose attributes are accurately determined.

[0156] Furthermore, the pattern accuracy is used to calculate the proportion of individuals who correctly judged the assessment attribute out of the total number of test individuals. The calculation formula is shown in the following relationship:

[0157]

[0158] Where: N p This indicates the number of people who correctly judged the assessment attributes.

[0159] Furthermore, this example generates 3000 participants for each simulation, assuming that all participants use the ideal response pattern as their observed response pattern. The simulation data generation process is as follows:

[0160] 1) Based on the subjects' potential knowledge level, the ideal response patterns were ranked according to their total scores, and the number of participants for each score was allocated according to the probability of a standard normal distribution. For cases where the knowledge levels differed but the ideal total scores were the same, an average allocation strategy was adopted.

[0161] 2) Use a probability of 0.3 for sliding, and repeat 15 times. The probability of sliding to different scores is distributed proportionally to the reciprocal of the score difference to ensure that the ideal score has a high probability of sliding to its surrounding values. As the difference between the observed score and the ideal score increases, the sliding probability decreases accordingly. A sliding probability matrix is ​​generated for each item.

[0162] 3) The numerical value of the observed response mode is determined by generating a random number U and comparing it with the probability value of the sliding probability matrix.

[0163] Furthermore, the test results are shown in Table 4 below:

[0164] Table 4. Test Results

[0165] frequency MMR PMR 1 0.953 0.948 2 0.958 0.992 3 0.956 0.971 4 0.961 0.956 5 0.969 0.968 6 0.976 0.954 7 0.963 0.942 8 0.943 0.962 9 0.977 0.994 10 0.973 0.985 11 0.977 0.993 12 0.959 0.992 13 0.977 0.950 14 0.990 0.969 15 0.967 0.989 average value 0.967 0.971 variance 0.000133 0.000324

[0166] Therefore, Table 4 shows that the above-mentioned method for evaluating and recommending virtual drills for underwater robot dam inspection is applicable to the virtual drill system for underwater robot dam inspection.

[0167] This invention also provides a computer device for performing the above-described... Figures 2 to 4 Recommended method for the virtual drill scenario of underwater robot dam inspection shown.

[0168] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 5As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 5 Take a processor 10 as an example.

[0169] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0170] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0171] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0172] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0173] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0174] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0175] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0176] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for recommending a virtual exercise scene for underwater robot dam body inspection, characterized in that, A method for underwater robot virtual exercise system, the method comprising: obtaining a real case attribute data set, an evaluation attribute set, and an exercise scene data set of a plurality of scenes to be exercised, the real case attribute data set and the exercise scene data set being stored in a scene attribute database in a structured manner and being associated with corresponding scenes; based on the evaluation attribute set, using the exercise scene data set to perform underwater robot dam inspection virtual exercise skill evaluation in the plurality of scenes to be exercised, to obtain a plurality of skill evaluation results; based on the plurality of skill evaluation results, predicting the evaluation of the underwater robot dam inspection virtual exercise skill in a plurality of initial unexercised scenes, to obtain a plurality of skill evaluation prediction results; based on the plurality of skill evaluation prediction results, determining a plurality of target unexercised scenes in the plurality of initial unexercised scenes; based on the real case attribute data set, calculating a plurality of scene importance values and a plurality of scene similarity values of the plurality of target unexercised scenes, respectively; based on the plurality of skill evaluation prediction results, the plurality of scene importance values, and the plurality of scene similarity values, determining a target recommended scene for underwater robot dam inspection virtual exercise.

2. The method of claim 1, wherein, based on the evaluation attribute set, using the exercise scene data set to perform underwater robot dam inspection virtual exercise skill evaluation in the plurality of scenes to be exercised, to obtain a plurality of skill evaluation results, comprising: determining a test matrix based on the evaluation attribute set; obtaining a first evaluation association matrix and a first score matrix of the scene to be exercised, the first evaluation association matrix being used to associate exercise evaluation items with exercise skill evaluation attributes, and the first score matrix being used to record the score of each evaluation item; based on the evaluation attribute set, the test matrix, the first evaluation association matrix, and the first score matrix, using the exercise scene data set to perform underwater robot dam inspection virtual exercise skill evaluation in the plurality of scenes to be exercised, to obtain the plurality of skill evaluation results.

3. The method of claim 2, wherein, determining a test matrix based on the evaluation attribute set, comprising: obtaining direct association relationships between each evaluation attribute in the evaluation attribute set and determining an adjacency matrix, the evaluation attributes being divided into first-level attributes and second-level attributes; obtaining comprehensive relationships between each evaluation attribute in the evaluation attribute set and determining a reachable matrix; obtaining the pass status of each second-level attribute in the evaluation attribute set and determining a cognitive structure; based on the adjacency matrix, the reachable matrix, and the cognitive structure, determining the test matrix.

4. The method of claim 1, wherein, based on the plurality of skill evaluation results, predicting the evaluation of the underwater robot dam inspection virtual exercise skill in a plurality of initial unexercised scenes, to obtain a plurality of skill evaluation prediction results, comprising: obtaining a second evaluation association matrix and a second score matrix of the plurality of initial unexercised scenes; Based on the multiple skill assessment evaluation results, the second assessment correlation matrix and the second score matrix, the underwater robot dam inspection virtual training skill assessment evaluation situation of the multiple initial untrained scenes is predicted, and the multiple skill assessment evaluation prediction results are obtained.

5. The method of claim 1, wherein, Based on the real case attribute data set, multiple scene importance degree values and multiple scene similarity values of the multiple target untrained scenes are calculated, including: Multiple attribute frequency weight values and multiple attribute similarity values of the real case attribute data set are calculated; Based on the multiple attribute frequency weight values, the multiple scene importance degree values are calculated; Based on expert experience and analytic hierarchy process, the multiple scene similarity values are calculated using the multiple attribute similarity values.

6. An underwater robotic virtual drill system, comprising: A system for performing the underwater robot dam inspection virtual training scene recommendation method as claimed in any one of claims 1 to 5; the system comprises: an underwater robot virtual training assessment evaluation module and a virtual training scene recommendation module.

7. A computer device, comprising: It comprises: A memory and a processor, which are connected in communication with each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the underwater robot dam inspection virtual training scene recommendation method of any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for making a computer execute the underwater robot dam inspection virtual training scene recommendation method of any one of claims 1 to 5.

9. A computer program product, characterised in that, It comprises computer instructions for making a computer execute the underwater robot dam inspection virtual training scene recommendation method of any one of claims 1 to 5.

Images