Multi-sensor feature selection and collaborative optimization method and device for unmanned system

Abstract

The invention discloses an unmanned system-oriented multi-sensor feature selection and collaborative optimization method and equipment. The method comprises the following steps of: acquiring data by using various sensors carried by an unmanned system; performing importance evaluation on the features in the initial data set by adopting a ReliefF algorithm, calculating the weight of each feature, and selecting important features; constructing a plurality of subtasks by using the important features and a non-dominated sorting algorithm; and carrying out collaborative optimization on the feature selection scheme in a self-adaptive manner through environment selection iteration, evaluating a feature subset, and finally obtaining an optimal feature selection scheme which is used for decision support of an unmanned system. According to the method, the problems of feature redundancy, high computing resource consumption and low optimization efficiency during multi-sensor data fusion of the unmanned system in a complex environment can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned systems, and in particular, proposes a multi-sensor feature selection and collaborative optimization method and device for unmanned systems. BACKGROUND

[0002] In recent years, unmanned systems have been widely applied in many fields, especially in autonomous navigation, environmental perception and intelligent decision-making. With the increasing demand for unmanned system applications, how to improve its performance and efficiency has become a hot research issue. Unmanned systems are usually equipped with multiple sensors, such as laser radar, camera, inertial measurement unit (IMU), temperature and humidity sensor, etc., to obtain environmental information and support decision-making. However, in the process of multi-sensor fusion, how to effectively select and optimize sensor features to improve the system's perception ability and decision-making accuracy still faces great challenges.

[0003] With the dramatic increase in the dimensionality of sensor data and the complexity of multi-task, traditional feature selection methods face the problems of high-dimensional data, redundant features and interdependence of data in unmanned systems. These problems not only increase the computational complexity, but also may lead to longer system response time, increased storage requirements, and even affect the real-time and accuracy of the system. Therefore, how to remove redundant information, select effective features, and use collaborative optimization methods to improve feature selection efficiency and system performance in the context of multi-sensor has become one of the key technologies for the development of unmanned systems.

[0004] Currently, feature selection methods for multi-sensor data mainly focus on feature extraction and selection based on single sensor data, or feature screening based on artificial rules. However, these methods often fail to adapt to different task requirements and environmental changes in complex and dynamic unmanned system tasks, and the computational efficiency and optimization effect are not ideal on high-dimensional data sets. Therefore, how to design adaptive and multi-task collaborative feature selection methods to improve the global optimization ability of the system has become a technical problem to be solved in the field of unmanned systems. SUMMARY

[0005] The present application proposes a multi-sensor feature selection and collaborative optimization method and device for unmanned systems to address the shortcomings of existing multi-sensor feature selection methods. Through this method, the time and computational resources required in the multi-sensor feature selection process can be effectively reduced. At the same time, through an adaptive mechanism, the most suitable auxiliary task feature selection scheme is dynamically selected according to the evolution state of the target task, so that the optimal feature subset is obtained in the optimization process.

[0006] The present application adopts the following technical solutions to solve the above technical problems: Step 1: Collect raw data using various sensors (such as vision sensors, radar, lidar, etc.) on the unmanned system, taking the actual environmental state collected by the sensors as a sample value; preprocess the collected raw data, including noise reduction, normalization, and other operations. Step 2: Preprocess the collected raw data and calculate the weight of each feature to select important features; these important features will serve as the initial feature selection scheme for subsequent subtask construction. Step 3: Construct multiple subtasks using the selected important features and non-dominated ranking algorithms; the construction of these subtasks is based on the importance of the features and the optimization objectives of the algorithm, providing a specific task framework for subsequent collaborative optimization. Step 4: Adaptively and collaboratively optimize the initial feature selection scheme through environmental selection iteration. During the iteration process, the optimization direction is continuously adjusted based on environmental feedback to finally obtain the optimal feature selection scheme. This scheme will provide accurate and efficient data support for the decision support of unmanned systems.

[0007] Furthermore, step two includes the following steps: Step 2.1: Use the ReliefF method to evaluate the importance of the original features, assessing the correlation between each feature and the class label to identify key features highly correlated with classification performance. For each feature... The weights are calculated as follows: (1) In equation (1), Features The weight, For the number of iterations, Indicates the index of the sample. Indicates from 1 to Quantity index; Indicates the category of the sample; This represents the number of nearest neighbor samples; It is a sample Class label; and Indicate category and categories proportion, Indicates sample and In features Differences on and They are samples of the same class; Indicates sample and In features Differences on and These are samples of a different class; Step 2.2: Based on the calculated feature weight values, sort all features in descending order to obtain the sorted feature set. and the corresponding weight set ; Step 2.3: Connect the points with the minimum and maximum feature weights to form a connecting line. Calculate the distance from each point in the feature to... The vertical distance is used to select the point with the largest distance as the knee point, and its feature weight is used as the threshold. ; Step 2.4: Retain weights greater than or equal to the threshold. The characteristics form an important feature set. This threshold is used for constructing subsequent low-dimensional tasks; features smaller than this threshold are considered redundant and are deleted. Furthermore, step three includes, Step 3.1: Obtain the set of important features Then, a population of appropriate size is randomly initialized to construct subtasks. Each individual in the population represents a subset of features, where each feature is either selected or not selected. Step 3.2: Multi-objective evolutionary optimization of the population in Step 3.1; Step 3.3: The final population obtained through evolution is divided into groups using non-dominated sorting. Layer, extract the optimal Pareto layer All feature subsets are The remaining layers The extracted feature subset is ; Step 3.4: Based on the results of the non-dominated sort, construct... Subtasks { , ,…, }. The first subtask Use directly As a subset of basic features. For the remaining subtasks. ( =2,3,…, ),Will and Combine to form sub-tasks The specific formula is as follows: ( =2,3,…, (2) Furthermore, step four includes, Step 4.1: Initialize parameters and population; Step 4.1.1 Initialize parameters; Current iteration number t=1, maximum iteration number ;initialization A table, in which all elements are initialized to 0; Step 4.1.2: Initialize the population, setting the population size to N; evenly distribute the N individuals in the population to each task. Corresponding to a subpopulation The subpopulation size is N / M; and ={ ,..., ,..., }, express The g-th individual in the dataset. Each individual is encoded using a real number. The position is ={ , ,...., ,..., },in Indicates the dimension of the dataset. Represents the g-th individual. The position of each feature, 0≤ ≤1, when When >0.6, it indicates that the first... One feature is selected and marked as '1', otherwise it is not selected and marked as '0'; Represents the g-th individual. The position of the last feature; Step 4.1.2: Calculate the fitness function value; The Middle The error rate for each individual is: (3) Equation (3) indicates the use of The error rate obtained when the feature selection scheme is used for classification is represented by c, where c is the number of classes in the dataset. This represents the proportion of correctly identified instances in class k; The Middle The number of features for each individual is: = (4) Equation (4) indicates the use of The number of features obtained when classifying using the feature selection scheme represented; The Middle The fitness function for each individual is: (5) Equation (5) indicates the use of The fitness function value obtained when classifying using the feature selection scheme represented by the example, where =0.999999, used to adjust the weight value between classification accuracy and the selected number of features; Using formulas (3), (4), and (5), select the N / M feature selection schemes with the smallest fitness function values ​​in each subpopulation as the t-th generation subpopulation for each task, and merge these N feature selection schemes as the t-th generation population to continue participating in evolution; Step 4.2: Calculate the similarity between tasks and select a set of candidate auxiliary tasks for the target task; Step 4.2.1, adopt Similarity calculations are performed between each pair of subtasks to construct a symmetric similarity matrix. The ( )th of the symmetric similarity matrix , ) elements represent and The structural similarity between elements is determined by using the median of all off-diagonal elements as the global similarity threshold. ; Step 4.2.2: Based on the threshold , build a Candidate auxiliary relation matrix This is used to identify auxiliary relationships between tasks. ≥ ,remember =1 The first in the matrix The first line The value of each element in the column is marked as 1, indicating a task. Can be used as Candidate auxiliary tasks; conversely, if < , =0 means The first in the matrix The first line The value of each element in the column is marked as 0; Step 4.2.3: For each target task From the matrix The Extract all values ​​of 1 from the row, and column index This constitutes its candidate auxiliary task set. ; Finally, each target task Each corresponds to a static candidate auxiliary task set. This provides an initial search space for subsequent adaptive auxiliary task selection mechanisms; Step 4.3: Based on the candidate auxiliary task set selected in Step 4.2, utilize... The reinforcement learning mechanism adaptively selects appropriate auxiliary tasks during the evolution of the target task; Step 4.3.1: Design for each target task The table contains three key components: state space construction, state partitioning mechanism, and action space definition. State space construction and introduction of migration survival rate And Elite Improvement Rate Two metrics, used to describe the global and local impacts of migration on the target population, are defined as follows: (6) In equation (6), where Indicates the previous generation of auxiliary population To the target population Offspring resulting from migration = ∩ This represents the number of offspring individuals that successfully survive in generation t. and Let them represent the target population in generation t-1 and generation t, respectively; (7) In equation (7), where and represent the fitness values ​​of the best individual in the target population in generation t-1 and generation t, respectively; The state division mechanism uses equations (6) and (7) to divide the current state of the target task into the following three categories: (Significant migration gain): When When ≥ 0.5; (Medium migration gain): When <0.5 and >0; (Poor migration performance): When MSR < 0.5 and = 0; The action space is defined, and according to step 4.2, it is limited to the set of candidate auxiliary tasks selected. ; Step 4.3.2: Based on 4.3.1, select the task-constructed [feature / method] for each feature. The table is used to select each objective task. The task corresponding to the maximum value is taken as the optimal auxiliary task, and then the subpopulation corresponding to the target task is evolved according to step 4.4. Step 4.3.3: Update according to reinforcement learning standards The table uses the following mechanism to... The value is updated iteratively: (8) In equation (8), where Current auxiliary task In state Instant rewards below, No. The state of the target population, In state The maximum of all possible auxiliary tasks value; ∈[0,1]: Learning rate, controlling the current reward pair The degree of impact of value updates; ∈[0,1]: Discount factor, used to measure the value of future rewards; in, Defined as: = (9); Step 4.4: Perform knowledge transfer. In each generation, for any target task... According to step 4.3.2, the current optimal auxiliary task is selected from the candidate set. Subsequently, the auxiliary task population was... The globally optimal particle participates in the target population. The update of failed particles is implemented using an improved CSO algorithm. The current iteration number is then set. With 0.5* In comparison, if <0.5* The population is evolved using formulas (10) and (12); if ≥0.5* The population is evolved using formulas (11) and (12): (10) (11) (12) Where r1, r2, r3 ∈ [0, 1] are random numbers; and These represent the current position information of the winning and losing particles in the target population of generation t, respectively. Indicates auxiliary population The global optimal particle position; Indicates auxiliary population The winning particle position in the middle, and These represent the velocities of the failed particles in generation t and generation (t+1), respectively. Step 4.5: After assigning t+1 to t, determine... > If the condition is met, the feature selection scheme with the minimum fitness function value is output for unmanned system recognition, thereby determining whether there are obstacles in the environment of the multi-sensor recognition of the unmanned system; otherwise, return to step four for execution.

[0008] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0009] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0010] Compared with existing technologies, the beneficial effects of this invention are reflected in: 1. This invention proposes a dimensionality reduction strategy for the multi-sensor feature selection problem in unmanned systems. The RelieF method is used to effectively filter the high-dimensional feature space, extracting the feature subset with the highest discriminative power. This strategy significantly reduces the feature dimensionality by evaluating the correlation and importance between features and eliminating redundant or irrelevant features. This provides a more concise and effective feature space for subsequent task generation, improves system processing efficiency, and lays a solid foundation for multi-sensor collaborative optimization.

[0011] 2. This invention designs an adaptive auxiliary task selection strategy, combining static task similarity analysis with a dynamic selection mechanism to automatically select the most suitable auxiliary task for the target task. This strategy can effectively reduce the risk of negative transfer, ensure efficient knowledge transfer between different tasks, thereby improving the overall accuracy and stability of task execution and facilitating subsequent collaborative optimization.

[0012] 3. This invention employs a collaborative optimization approach, utilizing an improved Cooperative Particle Swarm Optimization (CSO) algorithm for cross-task knowledge transfer. During the collaborative optimization process, adjusting the search direction and velocity of particles enhances inter-task collaboration, further improving the overall optimization effect. Combining multi-task learning and evolutionary mechanisms, this invention's collaborative optimization method efficiently shares and optimizes features between tasks, ultimately enabling adaptive decision-making and task execution for unmanned systems in complex environments. Attached Figure Description

[0013] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a schematic diagram illustrating the selection of a set of candidate auxiliary tasks for the target task according to the present invention; Figure 3 This invention is for A diagram illustrating the selection of the optimal auxiliary task when the target task is selected. Detailed Implementation

[0014] In this embodiment of the invention, a method and apparatus for multi-sensor feature selection and collaborative optimization for unmanned systems is provided. This involves preprocessing high-dimensional data collected by a multi-sensor unmanned system, selecting promising feature subsets for dimensionality reduction, and dividing the data into multiple feature selection sub-tasks for adaptive collaborative evolution. This improves the quality of the final optimal feature selection scheme set and reduces the time and space consumed in generating an adaptive multi-sensor feature selection scheme for the unmanned system to identify the category of collected objects. The method includes the following steps: Step 1: Use the sensors on the unmanned system to collect raw data and take the actual environmental state collected by the sensors as a sample value. Step 2: Preprocess the collected raw data and calculate the weight of each feature to select important features; these important features will serve as the initial feature selection scheme for subsequent subtask construction. Step 3: Construct multiple subtasks using the selected important features and non-dominated ranking algorithms; the construction of these subtasks is based on the importance of the features and the optimization objectives of the algorithm, providing a specific task framework for subsequent collaborative optimization. Step 4: Adaptively and collaboratively optimize the initial feature selection scheme through environmental selection iteration. During the iteration process, the optimization direction is continuously adjusted based on environmental feedback to finally obtain the optimal feature selection scheme. This scheme will provide accurate and efficient data support for the decision support of unmanned systems.

[0015] Reference Figure 1 As shown, the specific steps are as follows: Step 1: Collect raw data using various sensors (such as vision sensors, radar, lidar, etc.) onboard the unmanned system, using the actual environmental state captured by the sensors as a sample value. Preprocess the collected raw data, including denoising and normalization. Calculate the number of features in the denoised sample set and construct a feature set. Table 1 shows a feature set for unmanned system identification containing 9 statistical features and 3 samples. Features 1-3 are visual sensor features, features 4-6 are radar features, and features 7-9 are lidar features. Feature 1 is the mean brightness, 2 is the edge intensity, 3 is the texture complexity, 4 is the target distance, 5 is the target speed, 6 is the reflection intensity, 7 is the point cloud density, 8 is the height distribution, and 9 is the shape feature. Column 10 indicates the target category corresponding to the sample, where 0 represents non-obstacles (i.e., no obstacles in the environment) and 1 represents obstacles (i.e., obstacles exist in the environment). Table 1 Environmental State Characteristic Set

[0016] Step 2: Calculate the weight of each feature and select the important features; Step 2.1: Use the ReliefF method to evaluate the importance of the original features, assessing the correlation between each feature and the class label to identify key features highly correlated with classification performance. For each feature... The weights are calculated as follows: (1) In equation (1), Features The weight, For the number of iterations, Indicates the index of the sample. Indicates from 1 to Quantity index; Indicates the category of the sample; This represents the number of nearest neighbor samples; It is a sample Class label; and Indicate category and categories proportion, Indicates sample and In features Differences on and They are samples of the same class; Indicates sample and In features Differences on and These are samples from different classes; Table 2 shows the importance of each feature: Table 2 Feature Importance

[0017] Step 2.2: Feature Weight Sorting. Based on the calculated feature weight values, sort all features in descending order to obtain the sorted feature set. and the corresponding weight set ; Table 3 Feature sets sorted in descending order of feature weights

[0018] Step 2.3: Connect the points with the minimum and maximum feature weights to form a connecting line. Calculate the distance from each point in the remaining features to... The vertical distance is calculated, and the point with the largest distance is selected as the knee point (kneepoint=0.50), with its feature weight used as the threshold. ; Step 2.4: Retain weights greater than or equal to the threshold. The characteristics form an important feature set. , ={ , , , , , Features smaller than this threshold are used for constructing subsequent low-dimensional tasks; features smaller than this threshold are considered redundant and are removed. Table 4. Set of Important Features

[0019] Step 3: Construct multiple subtasks using these important features and non-dominated sorting algorithms; Step 3.1: Obtain the set of important features ={ , , , , , After that, a population of appropriate size is randomly initialized to construct subtasks (this population is only used to construct tasks and does not participate in subsequent evolution; in this experiment, the population size is 10). Each individual in the population represents a subset of features, where each feature is either selected or not selected. Step 3.2: Multi-objective evolutionary optimization of the population in Step 3.1; Step 3.3: The final population obtained through evolution is divided into groups using non-dominated sorting. Layer, in this experimental test =4, extract the optimal Pareto layer All feature subsets are The remaining layers ( The feature subset extracted from (=2,3,4) is ; The layers contain individuals {1, 0, 0, 0, 0, 0}, {0, 1, 0, 0, 0, 0}, and {1, 1, 0, 0, 0, 0}, so the union of all selected features is taken. The index is {1, 1, 0, 0, 0, 0, 0}; The layer contains individuals {1, 0, 0, 0, 0, 0} and {0, 0, 0, 0, 0, 1}, and the union of all features is taken. The index is {1, 0, 0, 0, 0, 0, 1}; The layers contain individuals of {1, 0, 0, 0, 0, 1} and {0, 0, 0, 0, 1, 1}. The index is {1, 0, 0, 0, 1, 1}; The layers contain individuals of the following values: {0, 0, 0, 0, 1, 0}, {0, 0, 1, 0, 0, 0, 0}, and {0, 1, 0, 0, 0, 0}. The index is {0, 1, 1, 0, 1, 0}; Step 3.4: Based on the results of the non-dominated sort, construct... Subtasks { , , , }. The first subtask Use directly As a subset of basic features. For the remaining subtasks. ( =2,3,4), will and Combine to form sub-tasks The specific formula is as follows: ( =2,3,4)(2) Based on steps 3.3 and 3.4, construct four sub-tasks. ={1, 1, 0, 0, 0, 0}, ={1, 1, 0, 0, 0, 1}, ={1, 1, 0, 0, 1, 1} and ={1, 1, 1, 0, 1, 0}; Step 4: Through environment selection iteration, the feature selection scheme is collaboratively optimized to obtain the optimal feature selection scheme for decision support of unmanned systems. Step 4.1: Initialize parameters and population; Step 4.1.1 Initialize parameters; Current iteration number t=1, maximum iteration number ;initialization A table, in which all elements are initialized to 0; Step 4.1.2: Initialize the population, setting the population size to N; evenly distribute the N individuals in the population to each task. Corresponding to a subpopulation The subpopulation size is N / 4 (i.e., N / M, M=4); and ={ ,..., ,..., }, express The g-th individual in the dataset. Each individual is encoded using a real number. The position is ={ , ,...., ,..., },in Indicates the dimension of the dataset. Represents the g-th individual. The position of each feature, 0≤ ≤1, when When >0.6, it indicates that the first... One feature is selected and marked as '1', otherwise it is not selected and marked as '0'; Represents the g-th individual. The position of the last feature; Step 4.1.2: Calculate the fitness function value; The Middle The error rate for each individual is: (3) Equation (3) indicates the use of The error rate obtained when the feature selection scheme is used for classification is represented by c, where c is the number of classes in the dataset. This represents the proportion of correctly identified instances in class k; The Middle The number of features for each individual is: = (4) Equation (4) indicates the use of The number of features obtained when classifying using the feature selection scheme represented; The Middle The fitness function for each individual is: (5) Equation (5) indicates the use of The fitness function value obtained when classifying using the feature selection scheme represented by the example, where =0.999999, used to adjust the weight value between classification accuracy and the selected number of features; Using formulas (3), (4), and (5), select the N / 4 feature selection schemes with the smallest fitness function values ​​in each subpopulation as the t-th generation subpopulation for each task, and merge these N feature selection schemes as the t-th generation population to continue participating in evolution; Step 4.2: Calculate the similarity between tasks and select a set of candidate auxiliary tasks for the target task (the four sub-tasks are used in turn for evolutionary optimization of the target task). Step 4.2.1, adopt Similarity calculations are performed between each pair of subtasks to construct a symmetric similarity matrix. The ( )th of the symmetric similarity matrix , ) elements represent and Structural similarity between them; such as Figure 2 As shown in (a), the similarity between tasks is calculated, and the task can be obtained from the third row of the table. With the task , and The similarity scores were 0.5, 0.25, and 0.4, respectively; the median of all off-diagonal elements was taken as the global similarity threshold. ( =0.29); Step 4.2.2: Based on the threshold , build a Candidate auxiliary relation matrix This is used to identify auxiliary relationships between tasks. ≥ ,remember =1 The first in the matrix The first line The value of each element in the column is marked as 1, indicating a task. Can be used as Candidate auxiliary tasks; conversely, if < , =0 means The first in the matrix The first line The value of each element in the column is marked as 0; for example Figure 2 As shown in (b), according to the third row of the table, 0.5 > 0.29, therefore... =1, and the other two candidate elements are calculated in the same way. =0, =0. ( Figure 2 The element marked as NULL in (a) Figure 2 The corresponding position of (b) is directly marked as 0. Step 4.2.3: For each target task From the matrix The Extract all values ​​of 1 from the row, and column index This constitutes its candidate auxiliary task set. ; Finally, each target task Each corresponds to a static candidate auxiliary task set. This provides an initial search space for the subsequent adaptive auxiliary task selection mechanism, resulting in a candidate auxiliary task set for each task. ={ }, ={ }, ={ , }and ={ , }; Step 4.3: Let the current evolutionary goal be... Based on the candidate auxiliary task set selected in step 4.2 ={ , },use Strengthen the learning mechanism, in During the evolution process, it adaptively selects appropriate auxiliary tasks; Step 4.3.1: For the current target task design The table contains three key components: state space construction, state partitioning mechanism, and action space definition. State space construction and introduction of migration survival rate And Elite Improvement Rate Two metrics, used to describe the global and local impacts of migration on the target population, are defined as follows: (6) In equation (6), where Indicates the previous generation of auxiliary population To the target population Offspring resulting from migration = ∩ This represents the number of offspring individuals that successfully survive in generation t. and Let them represent the target population in generation t-1 and generation t, respectively; (7) In equation (7), where and represent the fitness values ​​of the best individual in the target population in generation t-1 and generation t, respectively; The state division mechanism uses equations (6) and (7) to divide the current state of the target task into the following three categories: (Significant migration gain): When When ≥ 0.5; (Medium migration gain): When <0.5 and >0; (Poor migration performance): When MSR < 0.5 and = 0; Action space definition: Based on step 4.2, it is limited to the selected set of candidate auxiliary tasks. ={ , }, Therefore, mark and As the present The table's actions; Step 4.3.2: Based on 4.3.1, select the task-constructed [feature / method] for each feature. The table is used to select each objective task. The task corresponding to the maximum value is selected as the optimal auxiliary task, and then the subpopulation corresponding to the target task is evolved according to step 4.4; the current evolution task... population According to step 4.3.1 Table as follows Figure 3 As shown in (a); calculate each Values ​​such as Figure 3 As shown in (b), let The evolutionary state is According to row 2 of the table, 0.5 > 0.1, so we choose... As auxiliary tasks; Step 4.3.3: Update according to reinforcement learning standards The table uses the following mechanism to... The value is updated iteratively: (8) In equation (8), where Current auxiliary task In state Instant rewards below, No. The state of the target population, In state The maximum of all possible auxiliary tasks value; ∈[0,1]: Learning rate, controlling the current reward pair The degree of impact of value updates; ∈[0,1]: Discount factor, used to measure the value of future rewards; in, Defined as: = (9) choose As After completing the auxiliary task, evolve according to step 4.4. According to formulas (6) and (7), let After successful evolution, it enters the next state. Using equations (8) and (9), The value was updated from 0.5 to 0.587, and the update process is as follows: Figure 3 As shown in (c); Step 4.4: Perform knowledge transfer. In each generation, for the target task... According to step 4.3.2, the current optimal auxiliary task is selected from the candidate set. Subsequently, the auxiliary task population was... The globally optimal particle participates in the target population. The update of failed particles is implemented using an improved CSO algorithm. The current iteration number is then set. With 0.5* In comparison, if <0.5* The population is evolved using formulas (10) and (12); if ≥0.5* The population is evolved using formulas (11) and (12) (where i=3 and j=4): (10) (11) (12) Where r1, r2, r3 ∈ [0, 1] are random numbers; and These represent the current position information of the winning and losing particles in the target population of generation t, respectively. Indicates auxiliary population The global optimal particle position; Indicates auxiliary population The winning particle position in the middle, and These represent the velocities of the failed particles in generation t and generation (t+1), respectively. Step 4.5: After assigning t+1 to t, determine... > If the condition is met, the feature selection scheme with the minimum fitness function value is output for unmanned system recognition, thereby determining whether there are obstacles in the environment of the multi-sensor recognition of the unmanned system; otherwise, return to step four for execution.

[0020] To verify the effectiveness of the proposed solution, simulation experiments were conducted on 12 high-dimensional datasets. These datasets cover different feature dimensions and sample numbers; the purpose of the experiments was to optimize the size of the feature subset and improve classification accuracy using the feature selection method of the proposed solution. Dataset Description The following is basic information about 12 high-dimensional datasets: Table 5 High-dimensional dataset

[0021] Experimental results The feature selection method of this scheme was applied to each dataset, and the size of the feature subset and the classification accuracy were recorded. The following are detailed experimental results: Table 6 Experimental Results

[0022] In summary, this scheme, by optimizing feature selection, not only significantly reduces the size of the feature subset but also greatly improves classification accuracy, demonstrating high practical value and broad application prospects.

[0023] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0024] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0025] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0026] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0027] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0028] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0029] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for multi-sensor feature selection and collaborative optimization for unmanned systems, characterized in that, Follow these steps: Step 1: Use the sensors on the unmanned system to collect raw data and take the actual environmental state collected by the sensors as a sample value. Step 2: Preprocess the collected raw data and calculate the weight of each feature to filter out important features; These key features will serve as the initial feature selection scheme for subsequent subtask construction; Step 3: Construct multiple sub-tasks using the selected key features combined with a non-dominated sorting algorithm; The construction of these subtasks is based on the importance of features and the optimization objectives of the algorithm, providing a specific task framework for subsequent collaborative optimization; Step 4: Adaptively and collaboratively optimize the initial feature selection scheme through environmental selection iteration. During the iteration process, the optimization direction is continuously adjusted based on environmental feedback to finally obtain the optimal feature selection scheme. This scheme will provide accurate and efficient data support for the decision support of unmanned systems.

2. The multi-sensor feature selection and collaborative optimization method for unmanned systems according to claim 1, characterized in that: Step two includes, Step 2.1: Use the ReliefF method to evaluate the importance of the original features, assessing the correlation between each feature and the class label, thereby identifying key features highly correlated with classification performance; for each feature... The weights are calculated as follows: (1) In equation (1), Features The weight, For the number of iterations, Indicates the index of the sample. Indicates from 1 to Quantity index; Indicates the category of the sample; This represents the number of nearest neighbor samples; It is a sample Class label; and Indicate category and categories proportion, Indicates sample and In features Differences on and They are samples of the same class; Indicates sample and In features Differences on and These are samples of a different class; Step 2.2: Based on the calculated feature weight values, sort all features in descending order to obtain the sorted feature set. and the corresponding weight set ; Step 2.3: Connect the points with the minimum and maximum feature weights to form a connecting line. ; Calculate the connection line from each point in the feature. The vertical distance is used to select the point with the largest distance as the knee point, and the feature weight of the knee point is used as the threshold. ; Step 2.4: Retain weights greater than or equal to the threshold. The characteristics form an important feature set. This threshold is used for constructing subsequent low-dimensional tasks; features smaller than this threshold are considered redundant and are deleted.

3. The multi-sensor feature selection and collaborative optimization method for unmanned systems according to claim 2, characterized in that: Step three specifically includes, Step 3.1: Obtain the set of important features Then, a population of a set size is randomly initialized to construct subtasks. Each individual in the population represents a subset of features, where each feature is either selected or not selected. Step 3.2: Multi-objective evolutionary optimization of the population in Step 3.1; Step 3.3: The final population obtained through evolution is divided into groups using non-dominated sorting. Layer, extract the optimal Pareto layer All feature subsets are The remaining layers The extracted feature subset is ; Step 3.4: Based on the results of the non-dominated sort, construct... Subtasks { , ,…, }; among them, the first subtask Use directly As a subset of basic features; for the remaining subtasks , =2,3,…, ,Will and Combine to form sub-tasks The specific formula is as follows: ， =2,3,…, (2)。 4. The multi-sensor feature selection and cooperative optimization method for unmanned systems according to claim 3, characterized in that: Step four specifically includes, Step 4.1: Initialize parameters and population; Step 4.2: Calculate the similarity between tasks and select a set of candidate auxiliary tasks for the target task; Step 4.2.1, adopt Similarity calculations are performed between each pair of subtasks to construct a symmetric similarity matrix. The ( )th of the symmetric similarity matrix , ) elements represent and The structural similarity between elements is determined by using the median of all off-diagonal elements as the global similarity threshold. ; Step 4.2.2: Based on the threshold , build a Candidate auxiliary relation matrix Used to identify auxiliary relationships between tasks; if ≥ ,remember =1 The first in the matrix The first line The value of each element in the column is marked as 1, indicating a task. Can be used as Candidate auxiliary tasks; conversely, if < , =0 means The first in the matrix The first line The value of each element in the column is marked as 0; Step 4.2.3, for each target task From the matrix The Extract all values ​​of 1 from the row, and column index This constitutes its candidate auxiliary task set. ; Finally, each target task Each corresponds to a static candidate auxiliary task set. This provides an initial search space for subsequent adaptive auxiliary task selection mechanisms; Step 4.3: Based on the candidate auxiliary task set selected in Step 4.2, utilize... The reinforcement learning mechanism adaptively selects appropriate auxiliary tasks during the evolution of the target task; Step 4.3.1: Design for each target task The table contains three key components: state space construction, state partitioning mechanism, and action space definition. State space construction includes introducing migration survival rate. And Elite Improvement Rate Two metrics, used to describe the global and local impacts of migration on the target population, are defined as follows: (6) In equation (6), where Indicates the previous generation of auxiliary population To the target population Offspring resulting from migration = ∩ This represents the number of offspring individuals that successfully survive in generation t. and Let them represent the target population in generation t-1 and generation t, respectively; (7) In equation (7), where and represent the fitness values ​​of the best individual in the target population in generation t-1 and generation t, respectively; The state partitioning mechanism includes using equations (6) and (7) to divide the current state of the target task into the following three categories: Significant migration gain: when When ≥ 0.5; Medium migration gain: when < 0.5 and > 0 o'clock; Poor migration performance: When MSR < 0.5 and = 0; The action space definition includes, according to step 4.2, limiting it to the set of candidate auxiliary tasks selected from the pool. ; Step 4.3.2: Based on 4.3.1, select the task-constructed [feature / method] for each feature. The table is used to select each objective task. The task corresponding to the maximum value is taken as the optimal auxiliary task, and then the subpopulation corresponding to the target task is evolved according to step 4.

4. Step 4.3.3: Update according to reinforcement learning standards The table uses the following mechanism to... The value is updated iteratively: (8) In equation (8), where Indicates the current auxiliary task In state Instant rewards below, Indicates the first The state of the target population, In state The maximum of all possible auxiliary tasks value; ∈[0,1]: Learning rate, controlling the current reward pair The degree of impact of value updates; ∈[0,1]: Discount factor, used to measure the value of future rewards; in, Defined as: = (9) Step 4.4: Perform knowledge transfer; In each generation, for any target task According to step 4.3.2, the current optimal auxiliary task is selected from the candidate set. ; Subsequently, the auxiliary mission population The globally optimal particle participates in the target population. The update of failed particles is implemented using an improved CSO algorithm; the current iteration number is set. With 0.5* In comparison, if <0.5* The population is evolved using formulas (10) and (12); if ≥0.5* The population is evolved using formulas (11) and (12): (10) (11) (12) Where r1, r2, r3 ∈ [0, 1] are random numbers; and These represent the current position information of the winning and losing particles in the target population of generation t, respectively. Indicates auxiliary population The global optimal particle position; Indicates auxiliary population The winning particle position in the middle, and These represent the velocities of the failed particles in generation t and generation (t+1), respectively. Step 4.5: After assigning t+1 to t, determine... > If the condition is met, the feature selection scheme with the minimum fitness function value is output for unmanned system recognition, thereby determining whether there are obstacles in the environment of the multi-sensor recognition of the unmanned system; otherwise, return to step four for execution.

5. The multi-sensor feature selection and cooperative optimization method for unmanned systems according to claim 4, characterized in that: Step 4.1 specifically includes, Step 4.1.1 Initialize parameters; Current iteration number t=1, maximum iteration number ;initialization A table, in which all elements are initialized to 0; Step 4.1.2: Initialize the population, setting the population size to N; evenly distribute the N individuals in the population to each task. Corresponding to a subpopulation The subpopulation size is N / M; and ={ ,..., ,..., }, express The g-th individual in the dataset; each individual is encoded using a real number. The position is ={ , ,...., ,..., },in Indicates the dimension of the dataset. Represents the g-th individual. The position of each feature, 0≤ ≤1, when When >0.6, it indicates that the first... One feature is selected and marked as '1', otherwise it is not selected and marked as '0'; Represents the g-th individual. The position of the last feature; Step 4.1.3: Calculate the fitness function value; The Middle The error rate for each individual is: (3) Equation (3) indicates the use of The error rate obtained when the feature selection scheme is used for classification is represented by c, where c is the number of classes in the dataset. This represents the proportion of correctly identified instances in class k; The Middle The number of features for each individual is: = (4) Equation (4) indicates the use of The number of features obtained when classifying using the feature selection scheme represented; The Middle The fitness function for each individual is: （5） Equation (5) indicates the use of The fitness function value obtained when classifying using the feature selection scheme represented by the example, where =0.999999, used to adjust the weight value between classification accuracy and the selected number of features; Using formulas (3), (4), and (5), select the N / M feature selection schemes with the smallest fitness function values ​​in each subpopulation as the t-th generation subpopulation for each task, and merge these N feature selection schemes as the t-th generation subpopulation to continue participating in evolution.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the computer program is executed by the processor, it causes the processor to perform the steps of the method as described in any one of claims 1 to 5.

Images