Multi-agent cooperative agricultural navigation system control method without map and related device

Abstract

This application provides a control method and related equipment for a multi-agent collaborative agricultural mapless navigation system. The control method includes: acquiring voice information and acquiring a first image corresponding to the location indicated by the voice information; processing the first image to obtain computer vision results, an image feature map, and an attention heatmap; fusing the image feature map, the attention heatmap, and the computer vision results to obtain a second image containing multiple first regions; determining the risk level corresponding to each first region in the second image, and determining the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region; processing the travel strategy of each first region to obtain a travel path, and controlling the agricultural mapless navigation system to travel according to the travel path. This application improves the operating efficiency of agricultural machinery.

Description

Technical Field

[0001] This application relates to the field of intelligent agent technology, and in particular to a control method and related equipment for a multi-agent collaborative agricultural mapless navigation system. Background Technology

[0002] Unmanned agricultural machinery automatic navigation technology typically relies on key technologies such as environmental perception, high-precision positioning, and path planning. Through multi-source sensors including global navigation satellite systems, inertial navigation units, lidar, and machine vision, it enables agricultural machinery to automatically travel in straight lines, automatically align with other vehicles, and automatically turn in farmland. In recent years, extensive research has been conducted on high-precision farmland map construction, autonomous positioning and navigation of unmanned agricultural machinery, and full-scene path planning. For example, a high-precision farmland map construction method proposed for unmanned agricultural machinery models plot information layers and obstacle layers to meet the map accuracy requirements of automatic navigation and provide prior information for path planning and obstacle perception.

[0003] In one exemplary technology, multiple agricultural machines can work together in the same plot of land through the coordinated control of the master and slave machines, relying on high-precision positioning and pre-built operation routes for navigation control.

[0004] In another exemplary technology, by processing the operation trajectory or image data, a genetic algorithm is used to extract the navigation line for agricultural machinery operation, providing a more regular path for automatic navigation, which is suitable for parallel operation path planning in regular plot environments.

[0005] In another exemplary technology, the lateral and azimuth deviations between the current position of the agricultural machinery and the preset path are obtained in real time. Combined with the preset driving range and control algorithm, the agricultural machinery can automatically navigate on a given path to improve path tracking accuracy and operational stability.

[0006] In another exemplary technology, an agricultural inspection robot with autonomous walking, automatic operation and intelligent recognition capabilities is proposed. It can complete tasks such as inspection, growth data collection and pest identification in the field, which enhances the robot's adaptability and human-machine collaborative operation capabilities in complex field environments. However, its navigation method is still mainly based on single-machine perception and traditional path planning.

[0007] In another exemplary technique, graph neural networks and task graph modeling are used to divide farmland into multiple operational units. Through multi-agent collaborative task division and path planning, multiple agricultural robots can work collaboratively in complex farmland, improving operational efficiency and resource utilization. This type of technology emphasizes coordination and scheduling at the task level, but at the navigation level, it still largely relies on task graphs constructed from preprocessed environmental data or implicit map representations.

[0008] All of the above technologies require agricultural machinery to build maps for navigation. Map building requires the agricultural machinery to travel through the work area, which takes a long time and results in low operational efficiency. Summary of the Invention

[0009] Based on the above-mentioned technological status, this application provides a control method and related equipment for a multi-agent collaborative agricultural mapless navigation system to solve the problem of low efficiency in agricultural machinery operations.

[0010] To achieve the above-mentioned technical objectives, this application proposes the following technical solution:

[0011] Firstly, this application provides a control method for a multi-agent cooperative agricultural mapless navigation system, including:

[0012] Based on the voice intelligent agent in the agricultural mapless navigation system, voice information is acquired, and the first image corresponding to the location indicated by the voice information collected by the visual perception intelligent agent in the agricultural mapless navigation system is acquired.

[0013] The visual perception agent, feature analysis agent, and heatmap agent in the agricultural mapless navigation system are controlled to process the first image to obtain computer vision results, image feature maps, and attention heatmaps. The image feature maps contain semantic information of objects in the first image, the attention heatmaps contain regions of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image.

[0014] Based on the navigation decision-making agent in the agricultural mapless navigation system, the image feature map, the attention heatmap, and the computer vision results are fused to obtain a second image containing multiple first regions, where the first regions contain obstacles;

[0015] The obstacle avoidance agent in the agricultural mapless navigation system determines the risk level corresponding to each of the first regions in the second image, and determines the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, to detour, or to slow down the travel.

[0016] The path planning agent in the agricultural mapless navigation system processes the travel strategy of each of the first areas to obtain the travel path, and controls the agricultural mapless navigation system to travel according to the travel path.

[0017] In some implementations, determining the risk level corresponding to each of the first regions in the second image based on the obstacle avoidance agent in the agricultural mapless navigation system includes:

[0018] The obstacle avoidance agent is controlled to detect obstacles in the first area and obtain target parameters. The target parameters are used to indicate the size of the obstacle and the distance between the obstacle and the agricultural mapless navigation system.

[0019] Based on the target parameters, a risk analysis is performed on the obstacle to obtain the risk level corresponding to the first area.

[0020] In some implementations, determining the travel strategy of the agricultural mapless navigation system in the first area based on the risk level of the first area includes:

[0021] The marking color of the first area is determined according to the risk level;

[0022] When the marked color indicates that the first area is impassable, the travel strategy of the agricultural mapless navigation system in the first area is configured as impassable;

[0023] When the marked color indicates that the first area is to be detoured, the travel strategy of the agricultural mapless navigation system in the first area is configured to detour;

[0024] When the marked color indicates that the first area is for deceleration, the travel strategy of the agricultural mapless navigation system in the first area is configured to decelerate.

[0025] In some implementations, determining the travel strategy of the agricultural mapless navigation system in the first area based on the risk level of the first area includes:

[0026] Determine the terrain type of the first region;

[0027] Based on the risk level of the first area and the terrain type, the travel strategy of the agricultural mapless navigation system in the first area is determined.

[0028] In some embodiments, controlling the agricultural mapless navigation system to travel along the travel path includes:

[0029] The first area to be traversed by the agricultural mapless navigation system on the travel path is designated as the second area;

[0030] Based on the travel strategy of the agricultural mapless navigation system in the second area, the target operating parameters of the agricultural mapless navigation system are determined;

[0031] Based on the collaborative control and management intelligent agent in the agricultural mapless navigation system, the agricultural mapless navigation system is controlled to travel in the second area according to the target operating parameters.

[0032] In some embodiments, the step of controlling the agricultural mapless navigation system according to the target operating parameters by the collaborative control and management intelligent agent in the agricultural mapless navigation system includes:

[0033] Based on the aforementioned collaborative control management intelligent agent, control instructions are generated for each sub-parameter in the target operating parameters;

[0034] Based on the component corresponding to the sub-parameter in the agricultural mapless navigation system, the priority of the control command corresponding to the sub-parameter is determined;

[0035] According to the order of their respective priorities, the agricultural mapless navigation system is controlled to execute each of the control commands sequentially.

[0036] In some implementations, it also includes:

[0037] During the process of the agricultural mapless navigation system traveling along the travel path, the travel status of the agricultural mapless navigation system is acquired;

[0038] The travel performance parameters of the agricultural mapless navigation system are determined based on the travel status, and the travel performance parameters are stored.

[0039] Secondly, this application provides a mapless navigation system for agriculture, comprising:

[0040] The acquisition module is used to acquire voice information based on the voice agent in the agricultural mapless navigation system, and to acquire the first image corresponding to the position indicated by the voice information collected by the visual perception agent in the agricultural mapless navigation system.

[0041] The control module is used to control the visual perception agent, feature analysis agent, and heat map agent in the agricultural mapless navigation system to process the first image and obtain computer vision results, image feature maps, and attention heat maps. The image feature maps contain semantic information of objects in the first image, the attention heat maps contain regions of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image.

[0042] The fusion module is used to fuse the image feature map, the attention heatmap, and the computer vision results based on the navigation decision-making agent in the agricultural mapless navigation system to obtain a second image containing multiple first regions, wherein the first regions contain obstacles.

[0043] The determination module is used to determine the risk level corresponding to each first region in the second image based on the obstacle avoidance agent in the agricultural mapless navigation system, and to determine the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, detour, or slow down the travel.

[0044] The processing module is used to process the travel strategy of each of the first areas according to the path planning agent in the agricultural mapless navigation system, obtain the travel path, and control the agricultural mapless navigation system to travel according to the travel path.

[0045] Thirdly, this application provides an electronic device, including a memory and a processor, wherein,

[0046] The memory is connected to the processor and is used to store programs;

[0047] The processor is used to implement the multi-agent cooperative agricultural mapless navigation system control method as described in the first aspect or any implementation thereof by running the program in the memory.

[0048] Fourthly, this application provides a computer program product, which, when executed by a processor, implements the multi-agent cooperative agricultural mapless navigation system control method as described in the first aspect or any implementation thereof.

[0049] Fifthly, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the multi-agent cooperative agricultural mapless navigation system control method as described in the first aspect or any implementation thereof.

[0050] This application provides a multi-agent collaborative agricultural mapless navigation system control method and related equipment. Through multiple agents in the agricultural mapless navigation system, the system performs path planning without a map, thus realizing mapless path planning and avoiding the time-consuming process of creating a map, thereby improving the operational efficiency of agricultural machinery. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0052] Figure 1 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 1 .

[0053] Figure 2 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 2 .

[0054] Figure 3 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 3 .

[0055] Figure 4 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 4 .

[0056] Figure 5 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 5 .

[0057] Figure 6 This is a schematic diagram of the functional modules of an agricultural mapless navigation system provided in an embodiment of this application.

[0058] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] It should be noted that the user information (including but not limited to electrical equipment information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with relevant laws, regulations and standards, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0061] Unmanned agricultural machinery automatic navigation technology typically relies on key technologies such as environmental perception, high-precision positioning, and path planning. Through multi-source sensors including global navigation satellite systems, inertial navigation units, lidar, and machine vision, it enables agricultural machinery to automatically travel in straight lines, automatically align with other vehicles, and automatically turn in farmland. In recent years, extensive research has been conducted on high-precision farmland map construction, autonomous positioning and navigation of unmanned agricultural machinery, and full-scene path planning. For example, a high-precision farmland map construction method proposed for unmanned agricultural machinery models plot information layers and obstacle layers to meet the map accuracy requirements of automatic navigation and provide prior information for path planning and obstacle perception.

[0062] In one exemplary technology, multiple agricultural machines can work together in the same plot of land through the coordinated control of the master and slave machines, relying on high-precision positioning and pre-built operation routes for navigation control.

[0063] In another exemplary technology, by processing the operation trajectory or image data, a genetic algorithm is used to extract the navigation line for agricultural machinery operation, providing a more regular path for automatic navigation, which is suitable for parallel operation path planning in regular plot environments.

[0064] In another exemplary technology, the lateral and azimuth deviations between the current position of the agricultural machinery and the preset path are obtained in real time. Combined with the preset driving range and control algorithm, the agricultural machinery can automatically navigate on a given path to improve path tracking accuracy and operational stability.

[0065] In another exemplary technology, an agricultural inspection robot with autonomous walking, automatic operation and intelligent recognition capabilities is proposed. It can complete tasks such as inspection, growth data collection and pest identification in the field, which enhances the robot's adaptability and human-machine collaborative operation capabilities in complex field environments. However, its navigation method is still mainly based on single-machine perception and traditional path planning.

[0066] In another exemplary technique, graph neural networks and task graph modeling are used to divide farmland into multiple operational units. Through multi-agent collaborative task division and path planning, multiple agricultural robots can work collaboratively in complex farmland, improving operational efficiency and resource utilization. This type of technology emphasizes coordination and scheduling at the task level, but at the navigation level, it still largely relies on task graphs constructed from preprocessed environmental data or implicit map representations.

[0067] All of the above technologies require agricultural machinery to build maps for navigation. Map building requires the agricultural machinery to travel through the work area, which takes a long time and results in low operational efficiency.

[0068] In addition, the aforementioned exemplary technologies rely heavily on high-precision maps and rule-based operating environments, making them difficult to adapt to complex and ever-changing farmland scenarios.

[0069] Most of the aforementioned exemplary technologies implicitly rely on maps or semi-static assumptions, and are not true mapless navigation.

[0070] The robust perception and path planning capabilities in complex farmland environments are insufficient, and there is a heavy reliance on single intelligence and single sensor combinations.

[0071] Lacking the ability to conduct real-time collaborative planning for complex farmland scenarios, it is difficult to balance operational efficiency and safety.

[0072] To address the aforementioned technical problems, this application proposes a multi-agent cooperative agricultural mapless navigation system control method. The following detailed description of the multi-agent cooperative agricultural mapless navigation system control method is provided through various embodiments.

[0073] Reference Figure 1 , Figure 1 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 1 .like Figure 1 As shown, the multi-agent cooperative agricultural mapless navigation system control method provided in this embodiment includes:

[0074] Step S101: Acquire voice information based on the voice agent in the agricultural mapless navigation system, and acquire the first image corresponding to the location indicated by the voice information collected by the visual perception agent in the agricultural mapless navigation system.

[0075] In this embodiment, the executing entity is an agricultural mapless navigation system, which can be agricultural machinery or a drone used for farmland operations. For ease of description, the term "system" will be used to refer to the agricultural mapless navigation system below.

[0076] The system is configured with multiple layers of intelligent agents, each layer including but not limited to the following:

[0077] 1. Perceptual Intelligent Agent Layer:

[0078] The voice-activated agent, including a microphone array, acquires voice signals, performs noise reduction processing, and performs speech recognition, parsing user voice commands into navigation intentions. This agent possesses adaptive noise suppression and multilingual recognition capabilities.

[0079] The visual perception agent acquires environmental images via a camera, supporting high-resolution real-time image acquisition to provide a data source for subsequent visual analysis. This agent is responsible for image preprocessing, format conversion, and quality optimization.

[0080] The feature analysis agent, which can be a ViT (Vision Transformer) agent, can be a pre-trained model. This agent performs deep feature extraction and attention analysis on the input image, generating a color pseudo-color feature activation map and displaying in real-time the model's attention to different regions of the environment. This agent focuses on high-level semantic understanding and visual attention analysis.

[0081] A heatmap agent can be a VLM (Vision-Language Model) heatmap generating agent. Based on a visual-language model, the heatmap agent performs multimodal reasoning to generate a VLM attention heatmap, combining language understanding and visual analysis capabilities to achieve a higher level of scene understanding. This agent possesses cross-modal reasoning capabilities in both vision and language.

[0082] A temporal processing agent is used to perform temporal analysis on continuous image sequences, detect moving targets, track environmental changes, and provide temporal information support for dynamic decision-making. This agent is specifically designed for time-related tasks and motion prediction.

[0083] 2. Decision-making intelligent agent layer:

[0084] The navigation decision-making agent is the core of the system's decision-making process. It integrates multi-source perception information to perform environmental understanding, target analysis, and path planning. This agent adopts a hierarchical architecture, including four sub-modules: perception, planning, decision-making, and execution, and is responsible for formulating the overall navigation strategy.

[0085] The obstacle avoidance agent identifies obstacles, estimates distances, and assesses risks based on visual detection results, enabling local path replanning and emergency obstacle avoidance. This agent focuses on real-time safety analysis and emergency response.

[0086] The path planning agent combines global path information and local environmental perception to generate smooth and safe navigation paths, supporting dynamic multi-planning and multi-objective optimization. This agent is specifically designed to handle path calculation and trajectory optimization tasks.

[0087] 3. Control agent layer:

[0088] The collaborative control and management agent, acting as the system's decision-making and arbitration center, uniformly processes control signals from various sources, including voice commands, autonomous decisions, obstacle avoidance instructions, and manual input, achieving priority management and conflict resolution. This agent ensures coordinated operation among all agents.

[0089] The navigation execution agent translates high-level decisions into specific chassis control commands, including speed control, direction control, and motion planning, ensuring the robot's safe and stable operation. This agent is responsible for the precise translation of motion commands.

[0090] Chassis Control Agent: Directly controls the controllers and motors in the system to achieve precise hardware-level motion control, supporting both real chassis and simulated chassis modes. This agent focuses on low-level hardware interfaces and real-time control.

[0091] When the system starts running, it needs to be initialized. System initialization includes the following steps:

[0092] Hardware self-test checks the status of hardware devices such as cameras, microphones, IMU sensors, and encoders;

[0093] Model loading: Load the ViT model, VLM model, and navigation algorithm model in sequence.

[0094] Renderer initialization: Start the ViT feature map renderer, VLM heat map renderer, and real-time display system;

[0095] System integration testing verifies the integrity of communication and data processing links between modules;

[0096] Each agent initializes and enters a preparation state.

[0097] Enter the ready state and wait for the user to input voice.

[0098] After the system completes initialization, multimodal perception processing is performed. Specifically, the microphone array in the voice agent collects the user's voice commands, which are then converted into navigation commands for the system as voice information through noise reduction, ASR (Automatic Speech Recognition) recognition, and intent parsing.

[0099] The system then uses a visual perception agent to capture an image corresponding to the location indicated by the voice information; this image is defined as the first image. For example, the voice information contains navigation instructions, which specify the location the system needs to navigate to. Therefore, the system can control the visual perception agent to acquire an image of that location, thus obtaining the first image.

[0100] Step S102: Control the visual perception agent, feature analysis agent, and heat map agent in the agricultural mapless navigation system to process the first image and obtain computer vision results, image feature map, and attention heat map. The image feature map contains semantic information of objects in the first image, the attention heat map contains the regions of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image.

[0101] After obtaining the first image, the visual perception agent, the feature analysis agent, and the heatmap agent process the first image respectively. The visual perception agent is located in the CV (Computer Vision) channel. The visual perception agent performs edge detection and shape recognition on the first image within 30 to 50 milliseconds to obtain the computer vision result, which includes the shape recognition result and edge detection result of the objects in the first image.

[0102] The feature analysis agent processes the first image within 50 to 80 milliseconds to generate an image feature map. The image feature map contains semantic information about the objects in the first image, such as the shape, texture, and spatial relationships of the objects.

[0103] The heatmap agent generates an attention heatmap within 80 to 120 milliseconds. The attention heatmap contains the regions of interest from the first image. For example, if the region of interest is an area containing obstacles, then the redder or brighter areas in the attention heatmap represent the areas where obstacles exist.

[0104] Step S103: Based on the image feature map, attention heat map and computer vision results of the navigation decision agent in the agricultural mapless navigation system, a second image containing multiple first regions is obtained, and the first regions contain obstacles.

[0105] After obtaining the image feature map, attention heatmap, and computer vision results, the navigation decision agent in the system fuses the image feature map, attention heatmap, and computer vision results to obtain the environmental perception result, which is an image and is defined as the second image.

[0106] For example, CV (Computer Vision) is fast and accurate, but easily fooled by adversarial examples or changes in lighting (e.g., misidentifying a white truck as the sky); Vitality Interpretation (ViT) has a broad perspective and strong contextual understanding, but it is computationally intensive and may not be as detailed as CNN; Virtual Learning Model (VLM) has deep understanding and can interpret the meaning of "no entry," but it is slow to respond and may not accurately locate specific pixel coordinates. Therefore, by combining information obtained from these three methods, the image feature map, attention heatmap, and computer vision results are weighted and fused to obtain a second image with high-precision annotations. The second image includes bounding boxes to indicate the location of objects; colors to indicate the type of region; and labels on objects to indicate their actions and meanings.

[0107] The second image contains multiple labeled regions, which are defined as the first region, and the first region contains obstacles.

[0108] Step S104: Determine the risk level corresponding to each first region in the second image based on the obstacle avoidance agent in the agricultural mapless navigation system, and determine the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, detour, or slow down the travel.

[0109] After determining the second image, the obstacle avoidance agent determines the risk level corresponding to each first region in the second image. For example, the first regions are labeled with color, region type, etc., and the risk level is determined by the color and region type. For instance, risk levels may be Level 1, Level 2, and Level 3. If the color brightness is greater than a first brightness threshold and the region type is a fixed raised obstacle, the risk level is Level 1. If the color brightness is less than a second brightness threshold and the region type is a movable, lightweight obstacle (such as straw), the risk level is Level 3, where the second brightness threshold is less than the first brightness threshold. All other cases are determined to be at risk level 2.

[0110] After determining the risk level of the first area, the system determines its travel strategy in the first area based on the risk level. The travel strategy may be that the first area is impassable, the system should detour, or the system should slow down.

[0111] For example, if the risk level of the first area is level one, the system's travel strategy in the first area is to make it impassable; if the risk level of the first area is level two, the system's travel strategy in the first area is to detour; if the risk level of the first area is level three, the system's travel strategy in the first area is to slow down, that is, to move slowly.

[0112] Step S105: Process the travel strategy of each first area according to the path planning agent in the agricultural mapless navigation system to obtain the travel path, and control the agricultural mapless navigation system to travel according to the travel path.

[0113] After obtaining the travel strategy for each first region, the travel strategy for each first region can be processed by the path planning agent to obtain the travel path, thereby controlling the system to travel according to the travel path.

[0114] Furthermore, as the system travels along the path, its travel status is acquired, and based on this status, the system's travel performance parameters are determined and stored. The system can monitor its status during travel using these performance parameters, enabling performance tracking and error handling to ensure accurate and error-free operation.

[0115] In this embodiment, multiple intelligent agents in the agricultural mapless navigation system perform path planning, and the path is planned without a map, thus realizing mapless path planning for the agricultural mapless navigation system. This avoids the time-consuming process of creating a map and improves the operating efficiency of agricultural machinery.

[0116] Reference Figure 2 , Figure 2 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 2 ,based on Figure 1 In the embodiment shown, step S104 includes:

[0117] Step S201: Control the obstacle avoidance agent to detect obstacles in the first area and obtain target parameters. The target parameters are used to indicate the size of the obstacle and the distance between the obstacle and the agricultural mapless navigation system.

[0118] In this embodiment, an obstacle avoidance agent detects obstacles in the first region to obtain target parameters. These target parameters include, but are not limited to, the size of the obstacle and the distance between the obstacle and the system. The target parameters may also include the type of obstacle.

[0119] Step S202: Based on the target parameters, perform a risk analysis on the obstacles to obtain the risk level corresponding to the first area.

[0120] After determining the target parameters, a risk analysis of the obstacles is performed based on these parameters to obtain the risk level of the first area. For example, the larger the obstacle and the smaller the distance between the obstacle and the system, the greater the risk, and therefore the higher the risk level. Risk levels can be categorized as Level 1, Level 2, and Level 3, with Level 1 representing the highest risk; alternatively, risk levels can be marked with red, yellow, and green, with red indicating the highest risk and green indicating the lowest risk.

[0121] In this embodiment, by detecting obstacles, risk analysis is performed on the obstacles to accurately determine the risk level corresponding to the first area.

[0122] Figure 3 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 3 ,based on Figure 1 In the embodiment shown, step S104 includes:

[0123] Step S301: Determine the marking color of the first area based on the risk level.

[0124] In this embodiment, the marking color of the first area can be determined by the risk level. For example, if the risk level is level one, the marking color of the first area is red; if the risk level is level two, the marking color of the first area is yellow; and if the risk level is level three, the marking color of the first area is green.

[0125] Step S302: When the marking color indicates that the first area is impassable, the travel strategy of the agricultural mapless navigation system in the first area is configured to be impassable.

[0126] When the marker color is red, the first area is impassable. Therefore, the system's movement strategy in the first area is configured as impassable.

[0127] In step S303, when the marker color indicates that the first area is to be detoured, the travel strategy of the agricultural mapless navigation system in the first area is configured to detour.

[0128] When the marker color is yellow, the first area is a detour; therefore, the system's travel strategy in the first area is configured to detour.

[0129] Step S304: When the marker color indicates that the first area is for deceleration, the travel strategy of the agricultural mapless navigation system in the first area is configured to decelerate.

[0130] When the marker color is green, the first area is for slowing down, and the system's movement strategy in the first area is configured to slow down.

[0131] In this embodiment, the marking color of the first area is determined by the risk level, thereby accurately configuring the system with the travel strategy in the first area by using the marking color.

[0132] Figure 4 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 4 .based on Figure 1 or Figure 2 In the embodiment shown, step S104 includes:

[0133] Step S401: Determine the terrain type of the first area.

[0134] In this embodiment, the system determines the terrain type of the first area. The terrain type is, for example, flat land, steep slope, or depression.

[0135] Step S402: Determine the travel strategy of the agricultural mapless navigation system in the first area based on the risk level and terrain type of the first area.

[0136] After determining the terrain type of the first area, the system's movement strategy in the first area can be determined based on the risk level and terrain type.

[0137] For example, when the first risk level is Level 1, although Level 1 is impassable, the system faces fewer restrictions on movement on flat terrain. Therefore, the system's movement strategy in the first area can be detour. Detour, as mentioned in this application, refers to moving around obstacles. If the risk level is Level 2 or Level 3, Level 2 indicates detour, and Level 3 indicates slowing down. However, if the terrain is a steep slope or depression, there is a greater risk when the system moves. Therefore, the system's movement strategy in the first area can be configured as impassable. If the risk level is Level 3 and the terrain is flat, the system's movement strategy in the first area is to slow down.

[0138] In this embodiment, the system accurately determines the system's travel strategy in the first area based on the terrain type and risk level of the first area.

[0139] Figure 5 The flowchart of a multi-agent cooperative agricultural mapless navigation system control method provided in this application embodiment. Figure 5 .based on Figures 1 to 4 In any of the embodiments shown, step S105 includes:

[0140] Step S501: Determine the first area to be traversed on the travel path of the agricultural mapless navigation system, and designate it as the second area.

[0141] In this embodiment, the system travels along a path. When the system approaches a first area on the path, that first area is the first area the system is to travel to, and the first area is designated as the second area. The system approaching the first area means that the distance between the system and the first area is less than a preset distance.

[0142] Step S502: Determine the target operating parameters of the agricultural mapless navigation system based on the travel strategy of the agricultural mapless navigation system in the second area.

[0143] The system acquires its own travel strategy in the second region and determines the target operating parameters for the system in the second region based on this strategy. For example, if the travel strategy is to decelerate, the target operating parameter could be the rotational speed of the motor driving the system, which is less than a preset speed.

[0144] Step S503: Based on the collaborative control and management intelligent agent in the agricultural mapless navigation system, control the agricultural mapless navigation system to travel in the second area according to the target operating parameters.

[0145] After determining the target operating parameters, the system operates according to the target operating parameters through a collaborative control management agent, so as to travel in the second region.

[0146] Furthermore, the system can generate control commands corresponding to each sub-parameter of the target operating parameters through a collaborative control agent. The target operating parameters include sub-parameters such as speed, acceleration, and steering. Different sub-parameters have corresponding control components; for example, the control component for steering is the steering wheel, for acceleration it's the accelerator, and for speed it's the deceleration pedal. For each sub-parameter, a corresponding control command can be generated; for example, if the sub-parameter is acceleration, the control command is about the accelerator pedal opening. These control commands can be generated by a navigation execution agent, and the control commands generated by the navigation execution agent are the parameters that the corresponding sub-parameter's component can execute.

[0147] After determining the control commands, the priority of the control commands corresponding to the sub-parameters is determined based on the components they correspond to in the system. Specifically, the collaborative control management intelligence formulates relevant rules to prioritize and handle conflicts between control signals (control commands) from different sources, ensuring that each control command is processed in an orderly manner and minimizing or avoiding conflicts between signals. For example, steering has the highest priority, followed by acceleration, and speed has the lowest priority.

[0148] The system executes control commands sequentially according to their priority. For example, steering has the highest priority, followed by acceleration, and speed has the lowest priority. Therefore, the system first controls the steering according to the steering command, then adjusts the throttle opening according to the acceleration command, and finally adjusts the deceleration pedal opening based on the speed command.

[0149] Before the system executes each control command, each interface of the system is tested, and after the control command is executed, each piece of hardware (the component corresponding to the sub-parameter) of the system is monitored to ensure the stability of the system chassis operation.

[0150] In this embodiment, the system determines the target operating parameters of the system based on the travel strategy in the area to be traveled, so that the system travels in the corresponding area with the appropriate operating parameters.

[0151] Based on the above embodiments, the advantages of the multi-agent cooperative agricultural mapless navigation system control method of this application will be explained:

[0152] 1. An innovative three-channel parallel processing architecture of ViT+VLM+CV is proposed, with each channel focusing on different levels of visual analysis:

[0153] ViT channel: Based on high-level semantic understanding using deep learning, it generates attention feature maps;

[0154] VLM channel: Multimodal understanding ability combining visual-language reasoning;

[0155] CV channel: Provides fast, robust basic visual features.

[0156] Improve perception accuracy through multi-channel complementarity, reducing blind spots and misjudgments of single algorithms; enhance environmental adaptability, providing reliable perception capabilities under different lighting and weather conditions; ensure real-time performance by maximizing the use of computing resources through parallel processing, achieving smooth processing.

[0157] 2. Intelligent collaborative control and decision-making arbitration mechanism:

[0158] It solves the problems of "control conflicts, priority confusion, and security risks in multi-agent systems, and the difficulty of traditional single control modes in dealing with complex multi-source input scenarios."

[0159] Specifically, this is achieved by designing a hierarchical collaborative control manager:

[0160] Priority management: Emergency stop > Obstacle avoidance command > Voice command > Autonomous decision-making > Manual input;

[0161] Conflict resolution: A multi-dimensional arbitration algorithm based on timestamps, security levels, and task urgency;

[0162] Security Override: Security checks have the highest privileges under any circumstances, overriding all other control commands.

[0163] To ensure operational safety, a multi-layered safety inspection mechanism effectively prevents dangerous operations; to improve response efficiency, a clear priority system ensures rapid response in critical moments; and to enhance system robustness, the system can still safely degrade and operate even when some modules fail.

[0164] 3. Real-time generation and visualization of ViT feature maps based on intelligent agents:

[0165] This addresses the problem that the "black box" decision-making process of traditional navigation systems lacks interpretability, making it difficult for users to understand the system's perception priorities and decision-making basis.

[0166] This is achieved by integrating a dedicated ViT feature analysis agent. Real-time feature extraction generates a color pseudo-color feature activation map within 50-80ms for each image frame; attention visualization visually displays the model's attention to different areas of the environment; multiple display modes support overlay, side-by-side, and split-screen visualization; agent collaboration allows the ViT feature analysis agent to work collaboratively with other visual agents, providing multi-layered visual understanding.

[0167] Improving system transparency allows users to intuitively understand the system's key perception points; facilitating debugging and optimization allows developers to make targeted algorithm improvements based on feature maps; enhancing user trust, as the visualized decision-making process improves the system's acceptability; and enabling agent collaboration, with ViT agents working together with other agents to enhance overall perception capabilities.

[0168] 4. Agent-based temporal perception and dynamic environment adaptation:

[0169] This addresses the problem that "static visual analysis struggles to handle moving target prediction and temporal correlation analysis in dynamic environments."

[0170] The problem is solved by introducing a specialized temporal processing agent. Motion detection is performed based on inter-frame difference and optical flow algorithms to detect moving targets; trajectory prediction is achieved using Kalman filtering to predict the trajectory of dynamic targets; temporal association is established to create semantic associations between multiple frames, improving the continuity of environmental understanding; and agent collaboration is implemented, with the temporal processing agent collaborating in real time with the navigation decision-making agent and obstacle avoidance agent to share temporal information.

[0171] Enhanced dynamic environment adaptation effectively addresses moving obstacles and changing environmental conditions; improved prediction accuracy, with trajectory prediction based on historical data enhancing obstacle avoidance performance; optimized path planning, with temporal information providing richer decision-making basis for path planning; and enabled agent collaboration, with temporal agents providing key time-dimensional information to other agents.

[0172] 5. Adaptive performance optimization and resource management based on agent collaboration:

[0173] This addresses issues such as "high resource consumption, unstable performance, and difficulty in deployment of complex AI systems".

[0174] This problem is solved by an adaptive performance optimization mechanism based on agent collaboration. Dynamic resource allocation adjusts the computing resources of each agent based on task priority and load; model quantization supports lightweight deployment, reducing memory and computing requirements; automatic degradation automatically switches to a lightweight processing mode when resources are scarce; agent scheduling dynamically schedules the working status and priority of each agent according to system load. This lowers the deployment threshold, supports stable operation under different hardware configurations; improves system reliability, with the adaptive mechanism ensuring normal operation under various conditions; optimizes user experience, with smooth real-time performance guaranteeing a good interactive experience; and agent collaborative optimization achieves globally optimal resource allocation through cooperation between agents.

[0175] Corresponding to the above-mentioned multi-agent collaborative agricultural mapless navigation system control method, this application embodiment also provides an agricultural mapless navigation system. Figure 6This is a schematic diagram of a mapless agricultural navigation system provided in an embodiment of this application. The mapless agricultural navigation system 600 provided in this embodiment includes:

[0176] The acquisition module 610 is used to acquire voice information based on the voice agent in the agricultural mapless navigation system, and to acquire the first image corresponding to the location indicated by the voice information collected by the visual perception agent in the agricultural mapless navigation system.

[0177] The control module 620 is used to control the visual perception agent, feature analysis agent, and heat map agent in the agricultural mapless navigation system to process the first image and obtain computer vision results, image feature map, and attention heat map. The image feature map contains semantic information of objects in the first image, the attention heat map contains the region of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image.

[0178] The fusion module 630 is used to fuse the image feature map, attention heat map and computer vision results of the navigation decision-making agent in the agricultural mapless navigation system to obtain a second image containing multiple first regions, wherein the first regions contain obstacles.

[0179] The determination module 640 is used to determine the risk level corresponding to each first region in the second image based on the obstacle avoidance intelligent agent in the agricultural mapless navigation system, and to determine the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, detour, or slow down the travel.

[0180] The processing module 650 is used to process the travel strategy of each first area according to the path planning agent in the agricultural mapless navigation system, obtain the travel path, and control the agricultural mapless navigation system to travel according to the travel path.

[0181] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0182] The obstacle avoidance agent is controlled to detect obstacles in the first area and obtain target parameters. The target parameters are used to indicate the size of the obstacle and the distance between the obstacle and the agricultural mapless navigation system.

[0183] Based on the target parameters, a risk analysis is performed on the obstacles to obtain the risk level corresponding to the first area.

[0184] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0185] The color of the first zone is determined based on the risk level.

[0186] When the color marking indicates that the first area is impassable, the travel strategy of the agricultural mapless navigation system in the first area will be configured as impassable.

[0187] When the color-coded area indicates that the first area is to be detoured, the travel strategy of the agricultural mapless navigation system in the first area will be configured to detour.

[0188] When the color-coded indication of the first area is to slow down, the travel strategy of the agricultural mapless navigation system in the first area is configured to slow down.

[0189] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0190] Determine the terrain type of the first region;

[0191] Based on the risk level and terrain type of the first area, the travel strategy of the agricultural mapless navigation system in the first area is determined.

[0192] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0193] The first area to be traversed by the agricultural mapless navigation system along its travel path is designated as the second area.

[0194] Based on the travel strategy of the agricultural mapless navigation system in the second area, the target operating parameters of the agricultural mapless navigation system are determined;

[0195] Based on the collaborative control and management intelligent agent in the agricultural mapless navigation system, the agricultural mapless navigation system is controlled to travel in the second area according to the target operating parameters.

[0196] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0197] Based on the collaborative control management intelligent agent, control instructions are generated for each sub-parameter in the target operating parameters;

[0198] Based on the component corresponding to the sub-parameter in the agricultural mapless navigation system, determine the priority of the control command corresponding to the sub-parameter;

[0199] The agricultural mapless navigation system executes each control command sequentially according to its priority.

[0200] In some implementations, the agricultural mapless navigation system 600 is also used for:

[0201] During the process of the agricultural mapless navigation system traveling along the travel path, the travel status of the agricultural mapless navigation system is obtained;

[0202] The travel performance parameters of the agricultural mapless navigation system are determined based on the travel status, and the travel performance parameters are stored.

[0203] The agricultural mapless navigation system provided in the above embodiments of this application and the above-mentioned multi-agent collaborative agricultural mapless navigation system control method belong to the same application concept. They can execute the multi-agent collaborative agricultural mapless navigation system control method provided in any of the above embodiments of this application, and possess the corresponding functional modules and beneficial effects of executing the multi-agent collaborative agricultural mapless navigation system control method. Technical details not described in detail in this embodiment can be found in the specific processing content of the multi-agent collaborative agricultural mapless navigation system control method provided in the above embodiments of this application, and will not be repeated here.

[0204] The functions of each module in the agricultural mapless navigation system can be implemented by the same or different processors, and this application embodiment does not limit this.

[0205] It should be understood that the modules in the above-described agricultural mapless navigation system can be implemented by a processor calling firmware. For example, the system includes a processor connected to a memory containing instructions. The processor calls the instructions stored in the memory to implement any of the above methods or to implement the functions of each module of the device. The processor can be a general-purpose processor, such as a CPU or microprocessor, and the memory can be internal to the device or external to the system. Alternatively, the modules in the system can be implemented as hardware circuits. By designing the hardware circuits, some or all of the module functions can be implemented. The hardware circuits can be understood as one or more processors. For example, in one implementation, the hardware circuit is an ASIC, and the functions of some or all of the above modules are implemented by designing the logical relationships between the components within the circuit. In another implementation, the hardware circuit can be implemented using a PLD, such as an FPGA, which can include a large number of logic gates. The connection relationships between the logic gates are configured through a configuration file, thereby implementing the functions of some or all of the above modules. All modules of the above-described agricultural mapless navigation system can be implemented entirely by a processor calling firmware, entirely by hardware circuits, or partially by a processor calling firmware with the remaining parts implemented by hardware circuits.

[0206] In this application embodiment, a processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a CPU, microprocessor, GPU, or DSP. In another implementation, the processor can implement certain functions through the logical relationships of hardware circuits. These logical relationships are fixed or reconfigurable. For example, the processor may be a hardware circuit implemented as an ASIC or PLD, such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and configuring the hardware circuit can be understood as the processor loading instructions to implement the functions of some or all of the above modules. Furthermore, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as an NPU, TPU, or DPU.

[0207] As can be seen, each module in the above-mentioned agricultural mapless navigation system can be one or more processors (or processing circuits) configured to implement the above methods, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor types.

[0208] Furthermore, the modules in the above-mentioned mapless navigation system for agriculture can be integrated in whole or in part, or they can be implemented independently. In one implementation, these modules are integrated together and implemented in the form of a System-on-Chip (SoC). The SoC may include at least one processor for implementing any of the above methods or implementing the functions of the modules of the device. The at least one processor can be of different types, such as CPU and FPGA, CPU and artificial intelligence processor, CPU and GPU, etc.

[0209] This application provides a schematic diagram of the structure of an electronic device, see [link]. Figure 7 As shown, the electronic device includes a memory 700 and a processor 710; wherein the memory 700 is connected to the processor 710 and is used to store programs; the processor 710 is used to implement the multi-agent cooperative agricultural mapless navigation system control method disclosed in any of the above embodiments by running the programs stored in the memory 700.

[0210] Specifically, the aforementioned electronic device may further include: a bus, a communication interface 720, an input device 730, and an output device 740. The electronic device may also include a data transceiver module, an image monitoring module, and a signal monitoring module.

[0211] The processor 710, memory 700, communication interface 720, input device 730, and output device 740 are interconnected via a bus. Among them:

[0212] A bus can include a pathway for transmitting information between various components in an electronic device.

[0213] The processor 710 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0214] The processor 710 may include a main processor, as well as a baseband chip, modem, etc.

[0215] The memory 700 stores a program that executes the technical solution of this invention, and may also store an operating system and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory 700 may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0216] Input device 730 may include a device for receiving data and information input by a user, such as a keyboard, mouse, camera, scanner, light pen, voice input device, touch screen, pedometer, or gravity sensor.

[0217] Output device 740 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0218] The communication interface 720 may include a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0219] The processor 710 executes the program stored in the memory 700 and calls other devices, which can be used to implement each step of any of the multi-agent cooperative agricultural mapless navigation system control methods provided in the above embodiments of this application.

[0220] It should be noted that electronic devices can be in-vehicle terminals, mobile phones, wearable devices, or servers, etc.; or, they can include electronic devices such as in-vehicle terminals.

[0221] This application also proposes a chip, which includes a processor and a data interface. The processor reads and runs a program stored in the memory through the data interface to execute the multi-agent cooperative agricultural mapless navigation system control method described in any of the above embodiments. For the specific processing procedure and its beneficial effects, please refer to the above embodiments of the multi-agent cooperative agricultural mapless navigation system control method.

[0222] In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the multi-agent cooperative agricultural mapless navigation system control method according to various embodiments of this application as described in any of the above embodiments of this specification.

[0223] Computer program products can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the power device, as a standalone firmware package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0224] Furthermore, embodiments of this application may also be storage media storing computer programs, which are executed by a processor to perform the steps of the multi-agent cooperative agricultural mapless navigation system control method according to various embodiments of this application described in any of the above embodiments of this specification, specifically implementing the steps of the multi-agent cooperative agricultural mapless navigation system control method as described above.

[0225] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0226] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0227] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0228] The units of the apparatus in the various embodiments of this application can be merged, divided, and deleted according to actual needs.

[0229] It should be understood that the disclosed terminals, devices, and methods can be implemented in other ways, given the several embodiments provided in this application. For example, the terminal embodiments described above are merely illustrative. For instance, the division of modules or sub-modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple sub-modules or modules may be combined or integrated into another module, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0230] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0231] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or as firmware functional modules or sub-modules.

[0232] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer firmware, or a combination of both. To clearly illustrate the interchangeability of hardware and firmware, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or firmware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0233] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly using hardware, firmware units executed by a processor, or a combination of both. The firmware unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0234] Finally, 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 the element.

[0235] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A control method for a multi-agent cooperative agricultural mapless navigation system, characterized in that, include: Based on the voice intelligent agent in the agricultural mapless navigation system, voice information is acquired, and the first image corresponding to the location indicated by the voice information collected by the visual perception intelligent agent in the agricultural mapless navigation system is acquired. The visual perception agent, feature analysis agent, and heatmap agent in the agricultural mapless navigation system are controlled to process the first image to obtain computer vision results, image feature maps, and attention heatmaps. The image feature maps contain semantic information of objects in the first image, the attention heatmaps contain regions of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image. Based on the navigation decision-making agent in the agricultural mapless navigation system, the image feature map, the attention heatmap, and the computer vision results are fused to obtain a second image containing multiple first regions, where the first regions contain obstacles; The obstacle avoidance agent in the agricultural mapless navigation system determines the risk level corresponding to each of the first regions in the second image, and determines the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, to detour, or to slow down the travel. The path planning agent in the agricultural mapless navigation system processes the travel strategy of each of the first areas to obtain the travel path, and controls the agricultural mapless navigation system to travel according to the travel path.

2. The control method for a multi-agent cooperative agricultural mapless navigation system according to claim 1, characterized in that, The step of determining the risk level corresponding to each of the first regions in the second image based on the obstacle avoidance agent in the agricultural mapless navigation system includes: The obstacle avoidance agent is controlled to detect obstacles in the first area and obtain target parameters. The target parameters are used to indicate the size of the obstacle and the distance between the obstacle and the agricultural mapless navigation system. Based on the target parameters, a risk analysis is performed on the obstacle to obtain the risk level corresponding to the first area.

3. The control method for a multi-agent cooperative agricultural mapless navigation system according to claim 1, characterized in that, Determining the travel strategy of the agricultural mapless navigation system in the first area based on the risk level of the first area includes: The marking color of the first area is determined according to the risk level; When the marked color indicates that the first area is impassable, the travel strategy of the agricultural mapless navigation system in the first area is configured as impassable; When the marked color indicates that the first area is to be detoured, the travel strategy of the agricultural mapless navigation system in the first area is configured to detour; When the marked color indicates that the first area is for deceleration, the travel strategy of the agricultural mapless navigation system in the first area is configured to decelerate.

4. The control method for a multi-agent cooperative agricultural mapless navigation system according to claim 1, characterized in that, Determining the travel strategy of the agricultural mapless navigation system in the first area based on the risk level of the first area includes: Determine the terrain type of the first region; Based on the risk level of the first area and the terrain type, the travel strategy of the agricultural mapless navigation system in the first area is determined.

5. The control method for a multi-agent cooperative agricultural mapless navigation system according to claim 1, characterized in that, Controlling the agricultural mapless navigation system to travel along the travel path includes: The first area to be traversed by the agricultural mapless navigation system on the travel path is designated as the second area; Based on the travel strategy of the agricultural mapless navigation system in the second area, the target operating parameters of the agricultural mapless navigation system are determined; Based on the collaborative control and management intelligent agent in the agricultural mapless navigation system, the agricultural mapless navigation system is controlled to travel in the second area according to the target operating parameters.

6. The control method for a multi-agent cooperative agricultural mapless navigation system according to claim 5, characterized in that, The collaborative control and management intelligent agent based on the agricultural mapless navigation system controls the agricultural mapless navigation system to travel in the second area according to the target operating parameters, including: Based on the aforementioned collaborative control management intelligent agent, control instructions are generated for each sub-parameter in the target operating parameters; Based on the component corresponding to the sub-parameter in the agricultural mapless navigation system, the priority of the control command corresponding to the sub-parameter is determined; According to the order of their respective priorities, the agricultural mapless navigation system is controlled to execute each of the control commands sequentially.

7. The control method for a multi-agent cooperative agricultural mapless navigation system according to any one of claims 1-6, characterized in that, Also includes: During the process of the agricultural mapless navigation system traveling along the travel path, the travel status of the agricultural mapless navigation system is acquired; The travel performance parameters of the agricultural mapless navigation system are determined based on the travel status, and the travel performance parameters are stored.

8. An agricultural mapless navigation system, characterized in that, include: The acquisition module is used to acquire voice information based on the voice agent in the agricultural mapless navigation system, and to acquire the first image corresponding to the position indicated by the voice information collected by the visual perception agent in the agricultural mapless navigation system. The control module is used to control the visual perception agent, feature analysis agent, and heat map agent in the agricultural mapless navigation system to process the first image and obtain computer vision results, image feature maps, and attention heat maps. The image feature maps contain semantic information of objects in the first image, the attention heat maps contain regions of interest in the first image, and the computer vision results include shape recognition results and edge detection results of objects in the first image. The fusion module is used to fuse the image feature map, the attention heatmap, and the computer vision results based on the navigation decision-making agent in the agricultural mapless navigation system to obtain a second image containing multiple first regions, wherein the first regions contain obstacles. The determination module is used to determine the risk level corresponding to each first region in the second image based on the obstacle avoidance agent in the agricultural mapless navigation system, and to determine the travel strategy of the agricultural mapless navigation system in the first region based on the risk level of the first region. The travel strategy is used to indicate that the first region is impassable, detour, or slow down the travel. The processing module is used to process the travel strategy of each of the first areas according to the path planning agent in the agricultural mapless navigation system, obtain the travel path, and control the agricultural mapless navigation system to travel according to the travel path.

9. An electronic device, characterized in that, Including memory and processor, among which, The memory is connected to the processor and is used to store programs; The processor is used to implement the multi-agent cooperative agricultural mapless navigation system control method as described in any one of claims 1-7 by running the program in the memory.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the multi-agent cooperative agricultural mapless navigation system control method as described in any one of claims 1-7.

Images