CN-122008747-A - Amphibious robot system with cloud edge end cooperation and cooperation operation method thereof

CN122008747ACN 122008747 ACN122008747 ACN 122008747ACN-122008747-A

Abstract

The invention relates to the technical field of intelligent operation and maintenance of ocean pastures and robots, and discloses an amphibious robot system with cloud edge end cooperation and a cooperative operation method thereof. The system comprises an amphibious robot, an amphibious robot intelligent docking station and a cloud service platform comprising a center cloud and an edge cloud. The method comprises the steps of collecting and preprocessing data by a robot, transmitting the data back to a docking station, carrying out local identification and decision by the docking station, controlling the robot to execute a first physical operation, reporting a complex task to an edge cloud for enhancement processing, generating a cooperative instruction to control the robot to execute a second physical operation, and carrying out global model training and issuing by a central cloud. According to the invention, the cross-medium continuous autonomous operation of single equipment on the bank base of the marine pasture, on the water surface and under the water is realized, the problems of insufficient end-to-side calculation force and cloud dependency are solved through the cooperation of the cloud edge end, and the intelligent level and response efficiency of operation and maintenance are improved.

Inventors

HE HAOCHENG
LIN WENHUI
LIN KAIHONG
LIN MINGQUAN
LIN LIXIA
LI JUNYI
LIU CHANG
LU RENQUAN

Assignees

广东工业大学

Dates

Publication Date: 20260512
Application Date: 20260206

Claims (10)

1. An amphibious robot system with cooperative cloud edge ends, comprising: an amphibious robot (11) for performing inspection and work tasks in land, water and underwater environments and collecting environment and equipment status data; an amphibious robot intelligent docking station (12) connected with the amphibious robot (11) and used for automatically arranging and recovering the amphibious robot (11) and carrying out local preprocessing and intelligent analysis on data returned by the amphibious robot (11); The cloud service platform (13) is connected with the amphibious robot intelligent docking station (12), and the cloud service platform (13) comprises a center cloud platform (1301) and an edge cloud platform group (1302).
2. The cloud-edge collaborative amphibious robot system of claim 1, wherein the amphibious robot (11) comprises: an integrated wheel paddle (1101) for providing an amphibious scene power output; the environment monitoring module (1103) is provided with a multi-source sensor array and is used for collecting water quality, climate, images and video data; a pose control and navigation module (1104) for implementing autonomous navigation and motion control; A real-time communication module (1102) for exchanging data with the amphibious robot smart docking station (12); a power management module (1105) for managing power supply and energy consumption analysis of the amphibious robot (11); an expandable space (1106) is reserved for accessing the function expansion module.
3. The amphibious robot system with cooperative cloud edge and end according to claim 2, wherein the integrated wheel paddle (1101) adopts a wheel paddle leg composite mechanism and is driven by a brushless direct current motor so as to realize the switching between underwater propulsion and land walking.
4. The cloud-edge collaborative amphibious robot system of claim 1, wherein the amphibious robot smart dock (12) comprises: an amphibious robot lifting platform (1201) for automatically deploying and retrieving the amphibious robot (11); the power management module (1202) is used for charging the amphibious robot (11) and has an autonomous energy storage function; And the edge computing terminal (1203) is pre-deployed with a lightweight artificial intelligent model and is used for carrying out local preprocessing, feature extraction and intelligent grading judgment on data returned by the amphibious robot (11).
5. The cloud-edge collaborative amphibious robot system of claim 4, wherein the edge computing terminal (1203) is further configured to maintain basic intelligent decision-making functionality when communication is interrupted and upload processed structured data to the cloud service platform (13).
6. The cloud-edge-collaborative amphibious robot system according to claim 1, wherein the central cloud platform (1301) is configured to uniformly manage the edge Yun Pingtai group (1302), perform high-precision model training based on a global dataset, and issue an optimized model; the edge cloud platform in the edge Yun Pingtai group (1302) is deployed in a regional data center and is used for carrying out semantic enhancement and structuring processing on data streams uploaded by the amphibious robot intelligent docking station (12) in a management region.
7. An amphibious robot collaborative operation method based on the cloud edge end collaborative amphibious robot system of any one of claims 1-6, characterized by being cooperatively executed by an amphibious robot (11), an amphibious robot smart dock (12) and a cloud service platform (13), the method comprising the steps of: S1, acquiring original environment data and equipment pose data by the amphibious robot (11), generating reliable working data after filtering processing, and transmitting the reliable working data back to the amphibious robot intelligent docking station (12); S2, the amphibious robot intelligent docking station (12) calls a pre-deployed lightweight artificial intelligent model to identify the reliable working data, if the identification result meets a preset local event condition, a first control instruction is generated and issued to the amphibious robot (11), and a first physical operation action is executed by the amphibious robot intelligent docking station; s3, if the identification result meets the preset reporting event condition, forwarding the data to an edge cloud platform of the area where the data is located, carrying out semantic enhancement and structuring processing by the edge cloud platform, and if the identification result meets the global cooperative condition, generating a second control instruction and issuing the second control instruction, and executing a second physical operation action by the amphibious robot (11); and S4, training a high-precision model by the central cloud platform (1301) based on the global data set, and transmitting the optimized lightweight inference model to the edge cloud platform and the amphibious robot intelligent docking station (12) in an air downloading mode.
8. The method of claim 7, wherein in step S2, the preset local event condition includes at least one of identifying a damaged netting, exceeding a threshold of a water quality parameter, and monitoring abnormal gathering of fish shoal, and the first physical operation action includes approaching image acquisition by navigation to an event coordinate, or constant fixed point monitoring of an abnormal area by a sensor.
9. The method of claim 7, wherein in step S3, the preset global cooperative condition includes identifying at least one of a large algae outbreak, a contaminant spread across a plurality of cultivation areas, and a cage structure stress anomaly, and the second physical operation action includes performing a surrounding inspection in cooperation with other robots according to a cooperative operation instruction, or invoking a function expansion module installed additionally to the robot to perform a cleaning or auxiliary operation.
10. The method according to claim 7, wherein in step S4, the central cloud platform (1301) issues the lightweight inference model to the edge cloud platform and the amphibious robot smart dock (12) by means of over-the-air downloading.

Description

Amphibious robot system with cloud edge end cooperation and cooperation operation method thereof Technical Field The invention relates to the technical field of intelligent operation and maintenance of ocean pastures and robots, in particular to an amphibious robot system with cloud edge end cooperation and a cooperative operation method thereof. Background The marine pasture is used as a complex production system integrating ecological breeding, environment monitoring and facility maintenance, daily operation and maintenance of the complex production system need to span different spaces such as land-based land, water surface areas and seabed beds, and higher requirements are put on comprehensive capabilities such as operation continuity, medium-crossing maneuverability, quick response, intelligent decision-making, remote efficient coordination and the like of intelligent equipment so as to support long-term unmanned closed-loop autonomous operation and maintenance of the marine pasture. However, the existing marine pasture is generally monitored and operated by a special platform with single function, wherein an underwater robot can only execute constant-depth cruising and limited sensing and can not land or maneuver on the water surface, a water surface unmanned ship is limited to water surface inspection or net cage peripheral observation and does not have the potential capability, and shore-based mobile equipment is completely distributed outside a water area. In order to make up the capability gap, the partial scheme is to assemble multiple modules, namely, integrate a crawler belt or wheel type chassis, a propeller and a vertical channel propeller, and superimpose an independent computing unit or a shore-based GPU server for data processing. The method realizes preliminary automation in a local scene and replaces part of manual operation. However, such a solution still has the following prominent problems: 1. The amphibious autonomous operation capability is lacking, and the existing robot cannot realize seamless switching from shore-based departure, water surface inspection to underwater operation in the same task period. The pasture still needs to use a plurality of sets of equipment such as shore-based tracked vehicles, unmanned boats on the water surface, underwater robots and the like, and the scheduling and the connection depend on manpower, so that the operation and maintenance cost is greatly increased, and the task scheduling is complex and the cooperative efficiency is low due to the independent operation of a plurality of systems. 2. The cloud side end cooperative architecture is lacking, two modes of 'end side independent operation' or 'pure cloud processing' are adopted in the current system, wherein the end side is limited by the calculation power of an embedded platform and cannot deploy complex AI models such as fish shoal analysis and disease identification, the cloud processing has strong calculation power, but highly depends on a stable communication link, and the robot is extremely easy to lose energy due to network interruption in a marine complex environment. Therefore, the invention provides an amphibious robot system with cooperative cloud edge ends and a cooperative operation method thereof to solve the problems. Disclosure of Invention The invention aims to overcome the defects of the prior art and provides an amphibious robot system with cooperative cloud edge ends and a cooperative operation method thereof, so as to realize single-equipment and multi-task cross-medium autonomous intelligent operation and maintenance of an ocean pasture. In order to achieve the above purpose, the invention adopts the following technical scheme: in a first aspect, the present invention provides an amphibious robot system with cooperative cloud edge ends, including: The amphibious robot is used for executing inspection and operation tasks in land, water surface and underwater environments and collecting environment and equipment state data; The amphibious robot intelligent docking station is connected with the amphibious robot and used for automatically laying and recycling the amphibious robot and carrying out local preprocessing and intelligent analysis on data returned by the amphibious robot; the cloud service platform is connected with the amphibious robot intelligent docking station and comprises a center cloud platform and an edge cloud platform group. Further, the amphibious robot includes: the integrated wheel paddle is used for providing amphibious double-scene power output; the environment monitoring module is provided with a multi-source sensor array and is used for collecting water quality, climate, images and video data; the pose control and navigation module is used for realizing autonomous navigation and motion control; the real-time communication module is used for carrying out data exchange with the amphibious robot intelligent docking station; the power management module is