A modular edge computing control device, system, and networking method for multi-aircraft collaborative operations between air and ground.

Abstract

This invention discloses a modular edge computing control device, system, and networking method for multi-robot collaborative operations between air and ground robots. It relates to the field of multi-robot collaborative control. The invention addresses issues such as hardware incompatibility between air and ground robots, frequent power outages during battery swapping, sensor susceptibility to interference, mutual interference between control and image transmission links, and the lack of a unified processing mechanism for multi-node networking. The device includes a universal edge computing baseboard, a flight control or motion control expansion board, an external sensor board, a UPS power management board, and first and second communication link modules. The universal baseboard serves as a shared computing platform for both air and ground robots; the UPS power management board maintains power to the baseboard during main power failures or battery swapping; and the dual communication links transmit control commands and image data respectively. The invention also includes a global control master station and a collaborative operation system for air and ground robots, along with its networking method. This invention is applicable to multi-robot collaborative inspection, detection, emergency response, collaborative transportation, or relay communication scenarios.

Description

Technical Field

[0001] This invention relates to the fields of multi-robot collaborative control, edge computing devices, and heterogeneous communication networking, specifically to a modular edge computing control device, system, and networking method for multi-robot collaborative operations between air and ground. Background Technology

[0002] Multi-robot collaborative operation systems, consisting of aerial and ground robots, have broad application prospects in scenarios such as inspection and reconnaissance, emergency response, collaborative transportation, and relay communication. These systems typically require environmental perception, attitude and motion state calculation, low-level motion control, real-time image or video processing, and data interaction with a remote control master station to be completed on the robot itself. The real-time performance and reliability of these tasks largely depend on the comprehensive performance of the edge computing and control equipment. Therefore, the versatility, power supply continuity, communication reliability, and ease of maintenance of the edge computing and control equipment are key factors affecting the overall operating efficiency of the system.

[0003] Currently, aerial robots, such as rotary-wing UAVs and compound-wing aircraft, and ground robots, such as wheeled robots, tracked robots, and legged robots, typically employ control motherboards or computing control equipment designed separately for their respective platforms. Significant differences exist in the computing bases, control interfaces, and actuator interfaces between different platforms, resulting in a wide variety of spare parts and complex equipment replacement and maintenance operations. Particularly in multi-robot collaborative tasks requiring rapid deployment or long-term continuous operation, this heterogeneous hardware structure severely restricts the unified management of each node and the ability to quickly replace parts on-site, hindering the large-scale application of the system and the improvement of task execution efficiency.

[0004] Meanwhile, some existing devices lack a continuous power supply guarantee mechanism for edge computing bases. When the main power supply is disconnected, the power battery is replaced, or the power supply voltage fluctuates, edge computing nodes are prone to power loss and restart, leading to loss of node registration status, interruption of existing communication links, loss of context information for unfinished tasks, and a significant increase in system recovery costs. These problems are particularly prominent in scenarios requiring uninterrupted operation, such as emergency rescue and long-term inspection.

[0005] Furthermore, the measurement accuracy of sensors such as inertial measurement units, altitude detection units, and magnetic field detection units is easily affected by heat sources and electromagnetic interference in the environment. If these sensors are placed near high-performance processors or complex, high-speed signal areas, their output signal quality will significantly degrade. On the other hand, if the connection between sensors, flight control or motion control units, and edge computing units relies primarily on external wiring harnesses, it will increase the overall assembly complexity and adversely affect the long-term maintenance efficiency of the system.

[0006] In terms of communication, control commands, status feedback information, and heartbeat maintenance information are typically characterized by small data volume, stable transmission frequency, and high continuity requirements, while image or video data are characterized by large data volume, high transmission bandwidth requirements, and sensitivity to latency. If these two types of data are transmitted together using the same communication link, transmission congestion or link anomalies in image or video data can easily cause serious impacts on basic control and status interaction, and may even lead to system malfunction or task interruption.

[0007] In summary, existing technologies for aerial and ground robots suffer from issues such as incompatible hardware, frequent power outages during battery swapping, susceptibility to sensor interference, mutual interference between control and image transmission links, and a lack of unified handling mechanisms for node identity, platform type, and abnormal states during multi-node networking. Summary of the Invention

[0008] The purpose of this invention is to address the problems in existing technologies, such as hardware incompatibility between aerial and ground robots, frequent power outages during battery swapping, susceptibility to sensor interference, mutual interference between control and image transmission links, and the lack of a unified handling mechanism for node identity, platform type, and abnormal states during multi-node networking. Therefore, this invention provides a modular edge computing control device, system, and networking method for multi-robot collaborative operations between air and ground.

[0009] The technical solution of this invention is:

[0010] This invention provides a modular edge computing control device for multi-robot collaborative operations between air and ground, comprising a universal edge computing baseboard. This baseboard serves as a shared computing platform for both aerial and ground robots, and includes at least one of a processor unit, a storage unit, a camera interface, a high-speed data interface, a low-speed bus interface, and a display interface. These components are used to perform edge-side data processing, task scheduling, status management, image or video data processing, and data interaction with the global control master station. The universal edge computing baseboard also includes at least one board-to-board connector for stacked electrical connection with flight control / motion control expansion boards.

[0011] Furthermore, a flight control / motion control expansion board is also provided; the flight control / motion control expansion board is equipped with a microcontroller, actuator interface, debugging interface and sensor interface; the microcontroller is used to receive sensor data and generate actuator control signals; the actuator interface is used to adapt to motors, servos, wheel sets, tracks or other actuators.

[0012] Furthermore, it also includes an external sensor board, a UPS power management board, a first communication link module, and a second communication link module. The external sensor board is electrically connected to the flight control / motion control expansion board via an FPC flexible cable and is physically isolated from the general edge computing baseboard. The external sensor board includes at least one of an inertial measurement unit, an altitude detection unit, and a magnetic field detection unit. The UPS power management board is electrically connected to the general edge computing baseboard and includes a main power input terminal, a backup power interface, a charge / discharge management unit, a power path switching unit, and an output terminal, which supplies power to the general edge computing baseboard. The first communication link module and the second communication link module are respectively electrically connected to the general edge computing baseboard. The first communication link module is used to transmit control commands, status information, and heartbeat information, while the second communication link module is used to transmit image data or video data.

[0013] The present invention also provides an air-ground multi-robot collaborative operation robot system, including a global control master station, at least one air robot, at least one ground robot, and modular edge computing control devices as described above, respectively installed on the air robot or the ground robot.

[0014] Furthermore, the universal edge computing baseboard can be adapted to aerial or ground robots by replacing the flight control / motion control expansion board or loading different software parameters.

[0015] Furthermore, when the second communication link module malfunctions, the first communication link module continues to transmit control commands, status information, and heartbeat information, while the modular edge computing control device restricts or stops the transmission of image or video data.

[0016] This invention also provides a networking method for an air-to-ground multi-robot collaborative operation robot system. The air-to-ground multi-robot collaborative operation robot system includes a global control master station, at least one aerial robot, at least one ground robot, and modular edge computing control devices respectively installed on the aerial robot or the ground robot. The modular edge computing control devices are as described above. The method includes the following steps:

[0017] Step 1: Configure node identities for modular edge computing control devices deployed on aerial and ground robots;

[0018] Step 2: Load the aerial robot parameters or ground robot parameters according to the platform type corresponding to the node;

[0019] Step 3: The modular edge computing control device performs a self-test after power-on;

[0020] Step 4: After the self-test is completed, the modular edge computing control device registers its node with the global control master station through the first communication link module;

[0021] Step 5: After registration is completed, each node sends heartbeat information through the first communication link module at a preset cycle;

[0022] Step 6: Transmit control commands, status information, and heartbeat information through the first communication link module, and transmit image data or video data through the second communication link module;

[0023] Step 7: When the UPS power management board detects a main power supply abnormality, power replacement, or voltage fluctuation, the power path switching unit will switch the output to the backup power supply. The general edge computing baseboard will continue to operate and report the power supply abnormality status to the global control master station through the first communication link module.

[0024] Step 8: When the second communication link module malfunctions, the general edge computing baseboard stops, reduces the bit rate, or buffers image or video data, while continuing to maintain the transmission of control commands, status information, and heartbeat information through the first communication link module;

[0025] Step 9: The global control master station executes tasks such as task assignment, task pause, task resumption, node switching, or collaborative path adjustment based on the platform registration information, heartbeat information, power status, link status, and task status of each node.

[0026] Furthermore, the self-test described in step 3 includes:

[0027] Step 31: The general edge computing baseboard detects the status of the processor, storage, camera interface, first communication link module, and second communication link module;

[0028] Step 32: The flight controller / motion control expansion board checks the status of the microcontroller, actuator interface, sensor interface, and power detection interface;

[0029] Step 33: The UPS power management board detects the status of the main power supply, backup power supply, and output terminals;

[0030] Step 34: The external sensor board outputs the sensor's online status or calibration status.

[0031] Furthermore, the registration information mentioned in step 4 includes at least one of node identity, platform type, communication address, software version, sensor status, power status, and available link status.

[0032] Furthermore, the heartbeat information in step 5 includes at least one of the following: node online status, current task status, power status, location or motion status summary, and link status.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] 1. Regarding hardware compatibility issues

[0035] This invention employs a universal edge computing baseboard as a shared computing platform for both aerial and ground robots. It is electrically connected to flight control or motion control expansion boards via board-to-board connectors. The same universal baseboard can be used interchangeably between the two platforms. During field maintenance, only the expansion board needs to be replaced, rather than the entire computing unit, which helps reduce the types of spare parts and lowers maintenance complexity.

[0036] 2. Regarding the issue of power supply continuity

[0037] This invention integrates a UPS power management board into the device, which includes a main power input terminal, a backup power interface, a charge / discharge management unit, a power path switching unit, and an output terminal. When the main power supply fails, is replaced, or experiences voltage fluctuations, the power path switching unit switches the power supply path to the backup power supply, while the general-purpose edge computing baseboard continues to operate. Nodes are less prone to power loss and restart due to short-term power replacements or voltage fluctuations, and established registration status and communication links are maintained, which helps reduce task recovery costs.

[0038] 3. Regarding sensor interference issues

[0039] This invention electrically connects an external sensor board to a flight control or motion control expansion board via a flexible FPC cable, while physically isolating the sensor board from a universal edge computing baseboard, for example, through mounting posts, flexible connectors, or separate mounting positions. This reduces the impact of the universal edge computing baseboard on the sensor board's heat source and electromagnetic interference. Furthermore, using FPC cables helps reduce external wiring harnesses and lowers assembly complexity.

[0040] 4. Regarding the issue of mutual interference between communication links

[0041] This invention employs a first communication link module and a second communication link module for distributed transmission. When the second link malfunctions, the device restricts or halts the transmission of image and video data, while the first link continues to issue control commands and report status information.

[0042] 5. Regarding multi-node network management issues

[0043] This invention incorporates node identity configuration, self-checking, registration, and heartbeat maintenance processes in its networking method. Each edge computing control device performs a self-check upon power-up to confirm the normal operation of all modules, then sends registration information to the global control master station, which updates the node list. After registration, each node sends heartbeat information at a preset interval, and the master station determines whether a node is online based on the heartbeats. This mechanism facilitates the orderly management of the system when multiple nodes are connected. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the overall structure of a modular edge computing control device.

[0045] Figure 2 This is a schematic diagram of the stacked interconnection structure of a general edge computing baseboard and a flight control / motion control expansion board.

[0046] Figure 3 This is the power path and switching logic diagram of the UPS power management board.

[0047] Figure 4 This is a diagram showing the physical isolation and data acquisition structure of the external sensor board.

[0048] Figure 5 This is a diagram of the architecture of a multi-aircraft collaborative operation system between air and ground.

[0049] Figure 6 It is a data flow diagram that separates the control link and the image transmission link.

[0050] Figure 7 This is a flowchart of the networking method.

[0051] In the diagram: 100, Modular edge computing control device; 110, Universal edge computing baseboard; 111, Processor unit; 112, Camera interface; 113, High-speed data interface; 114, Low-speed bus interface; 120, Flight controller / motion control expansion board;

[0052] 121. Microcontroller; 122. Actuator Interface; 123. Power Detection Interface; 130. External Sensor Board; 131. Inertial Measurement Unit; 132. Altitude Detection Unit; 133. Magnetic Field Detection Unit; 140. UPS Power Management Board; 141. Main Power Input Terminal; 142. Backup Power Interface; 143. Charge / Discharge Management Unit; 144. Power Path Switching Unit; 145. Output Terminal; 150. First Communication Link Module; 160. Second Communication Link Module; 170. Board-to-Board Connector; 180. FPC Flexible Cable; 200. Global Control Master Station; 300. Aerial Robot; 400. Ground Robot. Detailed Implementation

[0053] Specific implementation method one: Combining Figures 1 to 6 This embodiment describes a modular edge computing control device equipped with a universal edge computing baseboard 110. The universal edge computing baseboard 110 serves as a shared computing platform for both the aerial robot 300 and the ground robot 400. It includes at least one of a processor unit 111, a storage unit, a camera interface 112, a high-speed data interface 113, a low-speed bus interface 114, and a display interface. These components are used to perform edge-side data processing, task scheduling, status management, image or video data processing, and data interaction with the global control master station. The universal edge computing baseboard 110 also includes at least one board-to-board connector 170 for stacked electrical connection with the flight control / motion control expansion board 120.

[0054] The universal edge computing baseboard 110 in this embodiment serves as the core carrier of the entire modular edge computing control device. It integrates a processor unit, a storage unit, a camera interface, a high-speed data interface, a low-speed bus interface, and a display interface. It can uniformly perform edge-side data processing, task scheduling, status management, image or video data processing, and data interaction tasks with the global control master station. Simultaneously, the board-to-board connectors on the baseboard provide a structural basis for stacked electrical connections with flight control or motion control expansion boards.

[0055] Specific Implementation Method Two: Combining Figures 1 to 6 This embodiment further includes a flight control / motion control expansion board 120; the flight control / motion control expansion board 120 includes a microcontroller 121, an actuator interface 122, a debugging interface, and a sensor interface; the microcontroller 121 is used to receive sensor data and generate actuator control signals; the actuator interface 122 is used to adapt to motors, servos, wheel sets, tracks, or other actuators.

[0056] With this configuration, the flight control / motion control expansion board in this embodiment is stacked with a general edge computing baseboard via a board-to-board connector, forming an upper and lower layer structure. The microcontroller integrated on the expansion board is responsible for receiving data from external sensors and generating control signals for the underlying actuators. The actuator interface is used to connect different actuators such as motors, servos, wheel sets, or tracks.

[0057] By replacing the flight control / motion control expansion board or loading different software parameters, the same universal edge computing baseboard can be adapted to either aerial or ground robots. When the flight control expansion board is installed and flight parameters are loaded, the device is used for drones; when the motion control expansion board is installed and ground motion parameters are loaded, the device is used for unmanned vehicles or tracked robots. This layered design decouples the computing base from the underlying control, solving the platform adaptation difficulties caused by the incompatibility of control and actuator interfaces in existing technologies. Furthermore, the board-to-board stacking connection reduces external wiring harnesses, improves integration, and enhances maintenance convenience.

[0058] Other components and connections are the same as in Specific Implementation Method 1.

[0059] Specific implementation method three: Combining Figures 1 to 6 This embodiment also includes an external sensor board 130, a UPS power management board 140, a first communication link module 150, and a second communication link module 160.

[0060] The external sensor board 130 is electrically connected to the flight control / motion control expansion board 120 via an FPC flexible cable 180 and is physically isolated from the general edge computing baseboard 110. The external sensor board 130 is provided with at least one of an inertial measurement unit 131, an altitude detection unit 132, and a magnetic field detection unit 133.

[0061] The UPS power management board 140 is electrically connected to the general edge computing baseboard 110. The UPS power management board 140 is provided with a main power input terminal 141, a backup power interface 142, a charge and discharge management unit 143, a power path switching unit 144, and an output terminal 145. The output terminal 145 supplies power to the general edge computing baseboard 110.

[0062] The first communication link module 150 and the second communication link module 160 are electrically connected to the general edge computing baseboard 110, respectively. The first communication link module 150 is used to transmit control commands, status information and heartbeat information, and the second communication link module 160 is used to transmit image data or video data.

[0063] In this configuration, the external sensor board is connected to the expansion board via an FPC cable and physically isolated from the general-purpose baseboard, reducing the impact of baseboard heat sources and electromagnetic interference on sensor acquisition accuracy. The UPS power management board switches to backup power when the main power supply fails or is replaced, maintaining continuous operation of the general-purpose baseboard and preventing loss of registration status due to node power failure and restart. Dual communication links are used for transmission: the first link transmits control commands, status, and heartbeats, while the second link transmits images and video, preventing congestion in image transmission from affecting basic control. Other components and connections are the same as in specific implementation methods one or two.

[0064] Specific implementation method four: Combination Figures 1 to 6 This embodiment describes a multi-robot collaborative operation robot system that includes a global control master station 200, at least one aerial robot 300, at least one ground robot 400, and modular edge computing control devices 100 as described above, respectively installed on the aerial robot 300 or the ground robot 400.

[0065] In this configuration, the global control master station in this embodiment is responsible for task distribution, node management, status monitoring, and collaborative scheduling. The aerial and ground robots, as execution ends, are each equipped with modular edge computing control devices. The edge computing control devices of each node interact with the master station via communication links. Furthermore, through this system architecture, the global control master station can uniformly manage the aerial and ground robots, achieving task collaboration and information sharing among multiple nodes, thus solving the problem of existing technologies where aerial and ground platforms are independent and lack a unified scheduling mechanism. Other components and connections are the same as in specific embodiments one, two, or three.

[0066] Specific Implementation Method Five: Combining Figures 1 to 6 This embodiment describes a universal edge computing baseboard 110 that can be adapted to an aerial robot 300 or a ground robot 400 by replacing the flight control / motion control expansion board 120 or loading different software parameters.

[0067] With this configuration, this implementation uses the same universal edge computing baseboard to achieve platform switching in two ways: one is to physically replace the flight control / motion control expansion board. With the above configuration, the same hardware baseboard can be flexibly switched between aerial and ground robots without the need to design separate computing motherboards for different platforms, further demonstrating the versatility and flexibility of the modular design of this invention.

[0068] Other components and connections are the same as in specific implementation methods one, two, three, or four.

[0069] Specific Implementation Method Six: Combination Figures 1 to 6 In this embodiment, when the second communication link module 160 malfunctions, the first communication link module 150 maintains the transmission of control commands, status information, and heartbeat information, while the modular edge computing control device 100 restricts or stops the transmission of image or video data.

[0070] With this configuration, if the second communication link malfunctions, the first communication link continues to transmit control and heartbeat information, while simultaneously limiting or stopping image transmission data to prevent image transmission problems from affecting basic control. Other components and connections are the same as in specific implementation methods one, two, three, four, or five.

[0071] Specific implementation method seven: Combination Figures 1 to 7 This embodiment describes a networking method in which the air-to-ground multi-robot collaborative operation robot system includes a global control master station 200, at least one aerial robot 300, at least one ground robot 400, and modular edge computing control devices 100 respectively installed on the aerial robot 300 or the ground robot 400, wherein the modular edge computing control devices 100 are as described above; the system is characterized by including the following steps:

[0072] Step 1: Configure node identities for the modular edge computing control device 100 deployed on the aerial robot 300 and the ground robot 400;

[0073] Step 2: Load the aerial robot parameters or ground robot parameters according to the platform type corresponding to the node;

[0074] Step 3: The modular edge computing control device 100 performs a self-test after being powered on;

[0075] Step 4: After the self-test is completed, the modular edge computing control device 100 registers the node with the global control master station 200 through the first communication link module 150;

[0076] Step 5: After registration is completed, each node sends heartbeat information through the first communication link module 150 at a preset period;

[0077] Step 6: Transmit control commands, status information and heartbeat information through the first communication link module 150, and transmit image data or video data through the second communication link module 160;

[0078] Step 7: When the UPS power management board 140 detects an abnormality in the main power supply, a power replacement, or a voltage fluctuation, the power path switching unit 144 switches the output terminal 145 to the backup power supply. The general edge computing baseboard 110 continues to operate and reports the abnormal power status to the global control master station 200 through the first communication link module 150.

[0079] Step 8: When the second communication link module 160 malfunctions, the general edge computing backplane 110 stops, reduces the bit rate, or buffers image or video data, and continues to maintain the transmission of control commands, status information, and heartbeat information through the first communication link module 150.

[0080] Step 9: The global control master station 200 executes tasks such as task assignment, task pause, task resumption, node switching, or collaborative path adjustment based on the platform registration information, heartbeat information, power status, link status, and task status of each node.

[0081] Specific implementation method eight: Combination Figures 1 to 7 This embodiment describes a self-test, specifically step 3, which includes:

[0082] Step 31: The general edge computing baseboard 110 detects the status of the processor, storage, camera interface, first communication link module 150 and second communication link module 160;

[0083] Step 32: The flight controller / motion control expansion board 120 detects the status of the microcontroller 121, actuator interface 122, sensor interface and power detection interface 123;

[0084] Step 33: The UPS power management board 140 detects the status of the main power supply, backup power supply and output terminal 145;

[0085] Step 34: The external sensor board 130 outputs the sensor's online status or calibration status.

[0086] This setup allows for layered self-testing to confirm the proper functioning of each module, preventing faulty modules from being deployed online. Other components and connections are the same as in implementation methods one, two, three, four, five, six, or seven.

[0087] Specific Implementation Method Nine: Combining Figures 1 to 7 This embodiment describes a registration information in step 4 that includes at least one of the following: node identity, platform type, communication address, software version, sensor status, power status, and available link status.

[0088] This configuration makes it easy to clearly define the specific content reported during node registration. Other components and connections are the same as any one of the specific implementation methods one through eight.

[0089] Specific Implementation Method Ten: Combining Figures 1 to 7 This embodiment describes a heartbeat information in step 5 that includes at least one of the following: node online status, current task status, power status, location or motion status summary, and link status.

[0090] With this setup, the node's online status confirms whether the device is still on the network; the current task status provides feedback on the progress of the ongoing task; the power status monitors whether the battery is sufficient; the position or motion status summary helps determine the robot's approximate location and movement trend; and the link status determines the connectivity of the first and second communication links. By periodically receiving these heartbeat messages, the master station can monitor the real-time survival status, task progress, remaining battery power, location information, and communication quality of each node. When a node experiences an anomaly (such as heartbeat timeout, low battery, or link interruption), the master station can promptly adjust the task scheduling. Other components and connections are the same as in any of the specific implementation methods one through nine.

[0091] Combination Figures 1 to 7 Description of embodiments of the present invention:

[0092] Example 1: Modular edge computing control device.

[0093] like Figure 1 As shown, the modular edge computing control device 100 in this embodiment includes a general edge computing baseboard 110, a flight control / motion control expansion board 120, an external sensor board 130, a UPS power management board 140, a first communication link module 150, and a second communication link module 160.

[0094] The universal edge computing base 110 serves as a shared computing platform for both the aerial robot 300 and the ground robot 400. The universal edge computing base 110 may include a processor unit 111, a storage unit, a camera interface 112, a high-speed data interface 113, a low-speed bus interface 114, and a display interface. The processor unit 111 is used to run edge computing programs, node management programs, vision processing programs, and communication management programs. In one embodiment, the processor unit 111 may employ a multi-core heterogeneous processor; preferably, it may employ an RK3588-type edge computing processor. The specific models described above are merely preferred implementations and do not constitute a limitation on the processor type.

[0095] like Figure 2 As shown, the flight control / motion control expansion board 120 is stacked on top of the general edge computing baseboard 110 via a board-to-board connector 170. The board-to-board connector 170 includes power terminals, low-speed bus terminals, and high-speed data terminals. The power terminals are used to transmit power or reference ground between the general edge computing baseboard 110 and the flight control / motion control expansion board 120; the low-speed bus terminals can carry UART, I2C, SPI, CAN, RS485, or other control buses; the high-speed data terminals can carry image, debugging, extended sensing, or high-speed peripheral data. These terminals are divided by area or pin group within the same board-to-board connector 170 to reduce external wiring and facilitate module replacement.

[0096] The flight control / motion control expansion board 120 includes a microcontroller 121, an actuator interface 122, a power detection interface 123, a debugging interface, and a sensor interface. The microcontroller 121 performs low-level real-time control, receives attitude, altitude, or magnetic field data from an external sensor board 130, and generates actuator control signals based on control commands from the general edge computing baseboard 110 or the global control master station 200. The actuator interface 122 may include a PWM interface, an ESC interface, a motor drive interface, a servo interface, a wheel assembly interface, a track interface, a relay interface, or other load control interfaces. In one embodiment, the microcontroller 121 may be an STM32H753-type microcontroller; this model is only a preferred embodiment.

[0097] like Figure 4As shown, the external sensor board 130 is connected to the flight control / motion control expansion board 120 via an FPC flexible cable 180. The external sensor board 130 is physically isolated from the general edge computing baseboard 110, which can be achieved through mounting posts, flexible connectors, separate mounting positions, or structural locations away from heat sources. The external sensor board 130 includes at least one of an inertial measurement unit 131, an altitude detection unit 132, and a magnetic field detection unit 133. The inertial measurement unit 131 can collect acceleration and angular velocity data, the altitude detection unit 132 can collect barometric altitude or distance data, and the magnetic field detection unit 133 can collect magnetic field direction or magnetic field strength data. In one embodiment, the inertial measurement unit 131 can be a BMI088 or ICM42688P type device, the magnetic field detection unit 133 can be an RM3100 type device, and the altitude detection unit 132 can be an ICP20100 type device; the above models are preferred embodiments.

[0098] The external sensor board 130 can also be equipped with mounting direction markings, positioning holes, notches, or silkscreen markings to ensure that the measurement coordinate system of the external sensor board 130 corresponds to the robot's body coordinate system. After assembly, the flight controller / motion control expansion board 120 can read the calibration parameters or stored parameters in the external sensor board 130 and use the sensing data for attitude estimation, motion state estimation, or closed-loop control of the actuator.

[0099] like Figure 3 As shown, the UPS power management board 140 includes a main power input terminal 141, a backup power interface 142, a charge / discharge management unit 143, a power path switching unit 144, and an output terminal 145. The main power input terminal 141 is used to connect to the robot's power battery, power bus, or external power supply; the backup power interface 142 is used to connect to a backup battery or energy storage unit; the charge / discharge management unit 143 is used to manage the charging, discharging, and status detection of the backup power supply; the power path switching unit 144 is used to switch the power supply path between the main power supply and the backup power supply; and the output terminal 145 is used to supply power to the general edge computing baseboard 110.

[0100] When the main power supply is normal, the UPS power management board 140 supplies power to the output terminal 145 through the main power input terminal 141, and can also charge or maintain the backup power supply. When the main power supply is abnormal, requires power replacement, or experiences voltage fluctuations, the power path switching unit 144 switches the power supply path of the output terminal 145 to the backup power supply based on the voltage detection result, ensuring that the general-purpose edge computing baseboard 110 remains powered. The UPS power management board 140 can also output the main power supply status, backup power supply status, switching status, or low battery status to the general-purpose edge computing baseboard 110 or the flight controller / motion control expansion board 120 through the power status reporting interface. In one embodiment, the charging and discharging management unit 143 may include a BQ25703 type charging management device, and the backup power protection or balancing part may include a BQ29209 type device; the above models are only preferred implementations.

[0101] The first communication link module 150 is used to transmit control commands, status information, and heartbeat information. The first communication link module 150 can be a low-speed long-range wireless communication module, a low-power wireless communication module, or a wired control communication module. In one embodiment, the first communication link module 150 can be a LoRa-type communication module. The second communication link module 160 is used to transmit image streams, video streams, or visual processing results. The second communication link module 160 can be a WiFi-type communication module, a high-speed data transmission module, an Ethernet module, or other high-bandwidth communication module. In one embodiment, the second communication link module 160 can cooperate with a low-latency image transmission scheme to transmit video data.

[0102] Through the above structure, the modular edge computing control device 100 forms a modular hardware structure in which computing, control, sensing, power supply, and communication work together. The general edge computing baseboard 110 provides unified computing capabilities, the flight control / motion control expansion board 120 provides platform-related low-level control capabilities, the external sensor board 130 provides physically isolated sensing and acquisition, the UPS power management board 140 provides power supply continuity, and the first communication link module 150 and the second communication link module 160 provide communication channels with separated data types.

[0103] Example 2: Multi-robot collaborative operation system between air and ground.

[0104] like Figure 5 As shown, the air-ground multi-robot collaborative operation robot system of this embodiment includes a global control master station 200, at least one air robot 300, at least one ground robot 400, and a modular edge computing control device 100 deployed on the air robot 300 and the ground robot 400 respectively.

[0105] The global control master station 200 is used for task distribution, node management, status monitoring, collaborative scheduling, and data aggregation. The global control master station 200 can be a ground station, a mobile control terminal, a portable computing device, or a server. The global control master station 200 maintains a node list, which includes at least the node's identity, platform type, communication address, heartbeat status, power status, image transmission link status, and task status.

[0106] The aerial robot 300 can be a rotary-wing UAV, a fixed-wing UAV, a compound-wing aircraft, or other aerial platform. The ground robot 400 can be a wheeled robot, a tracked robot, a legged robot, or other ground mobile platform. Both the aerial robot 300 and the ground robot 400 can use the same type of universal edge computing baseboard 110. For the aerial robot 300, the flight control / motion control expansion board 120 can connect to ESCs, motors, servos, or flight attitude control related interfaces and load aerial platform parameters; for the ground robot 400, the flight control / motion control expansion board 120 can connect to wheel sets, tracks, steering mechanisms, or mechanical actuators and load ground platform parameters.

[0107] Each modular edge computing control device 100 has a node identity. The node identity may include at least one of the following: node number, platform type, task role, communication address, and device key. The platform type is used to distinguish between aerial robot nodes and ground robot nodes. After loading the platform type parameters, the general edge computing baseboard 110 calls the corresponding control interface, sensor coordinate transformation parameters, motion constraint parameters, and communication configuration.

[0108] During system operation, the global control master station 200 sends task instructions, control instructions, or collaborative scheduling instructions to each node through the first communication link; each node also transmits its node status, motion status, power status, and heartbeat information back through the first communication link. The first communication link carries a relatively small amount of data and is mainly used to maintain system control and node online judgment.

[0109] The second communication link is used for transmitting image data, video data, or visual processing results. The aerial robot 300 can transmit overhead images or aerial videos via the second communication link, while the ground robot 400 can transmit close-up environmental images or operational perspective videos via the second communication link. The general-purpose edge computing baseboard 110 can acquire, encode, compress, stream, or extract target information from images or videos, and then transmit them to the global control master station 200 or other collaborative nodes via the second communication link.

[0110] When the main power supply of any node fails, is replaced, or experiences voltage fluctuations, the UPS power management board 140 of that node maintains power supply to the general-purpose edge computing baseboard 110. The general-purpose edge computing baseboard 110 can continue to maintain the node's registration status, heartbeat status, and task context, and report the power failure or switching status to the global control master station 200 via the first communication link. The global control master station 200 can adjust tasks according to this status, such as pausing high-power tasks of that node, reducing the image or video transmission bitrate, assigning other nodes to take over some tasks, or waiting for the main power supply to be restored.

[0111] When the second communication link malfunctions, the corresponding node stops or reduces image / video data transmission, while continuing to receive control commands, transmit status information, and send heartbeat information through the first communication link. The global control master station 200 can adjust its task strategy based on the image transmission link status, such as keeping the node moving at low speed, returning to base, waiting in place, switching observation nodes, or requesting the re-establishment of the second communication link. The above processing depends on the data division of labor between the first and second communication links and is not solely dependent on the specific communication standard.

[0112] Example 3: Network operation method.

[0113] like Figure 7 As shown, this embodiment provides a network operation method based on the above-mentioned air-ground multi-robot collaborative operation robot system.

[0114] Step S1, Node Identity Configuration. Configure node identities for the modular edge computing control device 100 deployed on the aerial robot 300 and the ground robot 400. Node identities can be written to the storage unit of the general edge computing baseboard 110, or assigned by the global control master station 200 upon initial access. Node identities are used to distinguish between different robot nodes.

[0115] Step S2, Platform Type Configuration. Based on the actual platform type installed on the node, load either aerial robot parameters or ground robot parameters. Aerial robot parameters may include flight control parameters, actuator mapping relationships, attitude sensor coordinate transformation relationships, and safety policies; ground robot parameters may include wheel or track control parameters, steering parameters, speed limit parameters, sensor coordinate transformation relationships, and actuator mapping relationships. The same universal edge computing baseboard 110 adapts to different platforms through different parameters and different flight control / motion control expansion boards 120.

[0116] Step S3, Power-on self-test. After the modular edge computing control device 100 is powered on, the general edge computing baseboard 110 detects the status of the processor, storage, camera interface, first communication link module 150 and second communication link module 160; the flight control / motion control expansion board 120 detects the status of the microcontroller 121, actuator interface 122, sensor interface and power detection interface 123; the UPS power management board 140 detects the status of the main power supply, backup power supply and output terminal 145; the external sensor board 130 outputs the sensor online status or calibration status.

[0117] Step S4, Node Registration. After self-testing, the modular edge computing control device 100 sends registration information to the global control master station 200 via the first communication link. Registration information may include node identity, platform type, communication address, software version, sensor status, power status, and available link status. Upon receiving the registration information, the global control master station 200 updates the node list and returns registration confirmation or configuration parameters to the nodes.

[0118] Step S5, Heartbeat Maintenance. After registration, each node sends heartbeat information through the first communication link at a preset period. The heartbeat information may include the node's online status, current task status, power status, location or motion status summary, and link status. The global control master station 200 determines whether a node is online based on the heartbeat information and updates the task scheduling strategy when the heartbeat times out or an anomaly occurs.

[0119] Step S6: Control and image transmission are separated. During system operation, control commands, status information, and heartbeat information are transmitted through the first communication link; image data, video data, or visual processing results are transmitted through the second communication link. The general-purpose edge computing baseboard 110 acquires, encodes, and transmits images or videos according to mission requirements, and the flight control / motion control expansion board 120 generates actuator control signals based on control commands and sensor data.

[0120] Step S7, Main Power Supply Anomaly Handling. When the UPS power management board 140 detects a main power supply anomaly, power replacement, or voltage fluctuation, the power path switching unit 144 switches the output terminal 145 to the backup power supply. The general edge computing backplane 110 continues to operate and reports the power supply anomaly status to the global control master station 200 via the first communication link. The global control master station 200 can adjust the node's tasks according to the backup power status, and the node can also reduce image / video transmission load or limit high-power execution actions according to local policies.

[0121] Step S8, image transmission link anomaly handling. When the second communication link fails, the general edge computing baseboard 110 stops, reduces the bit rate, or buffers image / video data, while continuing to transmit control commands, status information, and heartbeat information through the first communication link. The global control master station 200 decides whether to switch the video source, adjust the node's task, or request the re-establishment of the second communication link based on the status of the second communication link.

[0122] Step S9, Master Station Scheduling. The global control master station 200 executes task assignment, task pause, task resumption, node switching, or collaborative path adjustment based on the platform registration information, heartbeat information, power status, link status, and task status of each node. Through the above methods, the system operates in a multi-node air-to-ground network with the support of general hardware base, dual-link communication, and UPS power supply.

[0123] This invention boasts a high degree of versatility and low maintenance costs. It employs a modular architecture of a "universal edge computing baseboard + customized expansion board," enabling aerial and ground robots to share the same computing platform. By loading different software parameters, control programs, or connecting different actuators, node roles can be switched, thereby reducing the types of spare parts, lowering maintenance and replacement costs, and improving on-site support efficiency.

[0124] This invention offers excellent power supply continuity and strong continuous operation capability. By incorporating a UPS power management board, it provides uninterrupted power to edge computing devices. When the power battery is replaced, the main power supply is disconnected, or voltage fluctuations occur, it can maintain the continuous operation of the core computing unit, reducing node power failures and restarts, communication interruptions, and network reconstruction, thereby improving the system's continuous operation capability.

[0125] This invention features high communication reliability and low image transmission latency. It employs a communication strategy that separates control data from visual data, enabling low-dimensional control commands and status information to be transmitted via a highly reliable link, while high-dimensional video data is transmitted via a high-bandwidth link. This balances the reliability of the control link in complex environments with the low latency of the visual link, thereby enhancing real-time interaction capabilities during air-to-ground collaborative operations.

[0126] This invention exhibits good sensing stability and high measurement accuracy. By physically isolating the external sensor board from the high-performance general-purpose edge computing baseboard, this invention helps reduce the impact of motherboard heat generation and electromagnetic interference on the inertial measurement unit and altitude detection element, thereby improving the stability and accuracy of attitude calculation, motion state detection, and environmental perception.

[0127] This invention features high system integration and strong scalability. Through board-to-board connections, modular stacking, and a separate sensor design, it improves hardware integration and reduces the installation and maintenance complexity caused by external wiring harnesses. Furthermore, this architecture facilitates the expansion of communication, sensing, or execution modules according to different task requirements, making it suitable for various air-ground collaborative operation scenarios.

[0128] This invention is suitable for multi-agent collaborative applications in complex environments. It not only meets the unified deployment requirements of heterogeneous platforms for aerial and ground robots, but also maintains good communication, power supply, and control stability in confined spaces and other complex environments. Therefore, it is suitable for various multi-agent collaborative tasks such as inspection and reconnaissance, emergency rescue, hazardous operations, and environmental detection.

[0129] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make other changes within the spirit of the invention and apply it to fields not mentioned in the invention. Of course, all such changes made in accordance with the spirit of the invention should be included within the scope of protection claimed by the invention.

Claims

1. A modular edge computing control device for multi-machine collaborative operation between air and ground, characterized in that: It is equipped with a universal edge computing baseboard (110). The universal edge computing base (110) serves as a computing base shared by the aerial robot (300) and the ground robot (400). It is equipped with at least one of the following: a processor unit (111), a storage unit, a camera interface (112), a high-speed data interface (113), a low-speed bus interface (114), and a display interface. It is used to perform edge-side data processing, task scheduling, status management, image or video data processing, and data interaction with the global control master station. The general edge computing baseboard (110) is also provided with at least one board-to-board connector (170) for stacking electrical connection with the flight control / motion control expansion board (120).

2. The modular edge computing control device for multi-machine collaborative operation between air and ground as described in claim 1, characterized in that: It also features a flight control / motion control expansion board (120); The flight control / motion control expansion board (120) is equipped with a microcontroller (121), an actuator interface (122), a debugging interface, and a sensor interface; The microcontroller (121) is used to receive sensor data and generate actuator control signals; The actuator interface (122) is used to adapt to motors, servos, wheel sets, tracks or other actuators.

3. A modular edge computing control device for multi-machine collaborative operation between air and ground as described in claim 2, characterized in that: It is also equipped with an external sensor board (130), a UPS power management board (140), a first communication link module (150), and a second communication link module (160). The external sensor board (130) is electrically connected to the flight control / motion control expansion board (120) via an FPC flexible cable (180) and is physically isolated from the general edge computing baseboard (110). The external sensor board (130) is provided with at least one of an inertial measurement unit (131), an altitude detection unit (132), and a magnetic field detection unit (133). The UPS power management board (140) is electrically connected to the general edge computing baseboard (110). The UPS power management board (140) is provided with a main power input terminal (141), a backup power interface (142), a charge and discharge management unit (143), a power path switching unit (144), and an output terminal (145). The output terminal (145) supplies power to the general edge computing baseboard (110). The first communication link module (150) and the second communication link module (160) are electrically connected to the general edge computing baseboard (110), respectively. The first communication link module (150) is used to transmit control commands, status information and heartbeat information, and the second communication link module (160) is used to transmit image data or video data.

4. A multi-robot collaborative operation system for air and ground operations, characterized in that: It includes a global control master station (200), at least one aerial robot (300), at least one ground robot (400), and a modular edge computing control device (100) as described in claim 3, respectively disposed on the aerial robot (300) or the ground robot (400).

5. The multi-robot collaborative operation system for air and ground operations according to claim 4, characterized in that: The universal edge computing baseboard (110) can be adapted to the aerial robot (300) or the ground robot (400) by replacing the flight control / motion control expansion board (120) or loading different software parameters.

6. The multi-robot collaborative operation system for air and ground operations according to claim 5, characterized in that: When the second communication link module (160) malfunctions, the first communication link module (150) continues to transmit control commands, status information and heartbeat information, and the modular edge computing control device (100) restricts or stops the transmission of image or video data.

7. A networking method for an air-ground multi-robot collaborative operation robot system, characterized in that, The air-ground multi-robot collaborative operation robot system includes a global control master station (200), at least one aerial robot (300), at least one ground robot (400), and modular edge computing control devices (100) respectively installed on the aerial robot (300) or the ground robot (400), wherein the modular edge computing control devices (100) are as described in any one of claims 1 to 3; the networking method includes the following steps: Step 1: Configure node identities for the modular edge computing control devices (100) deployed on the aerial robot (300) and the ground robot (400); Step 2: Load the aerial robot parameters or ground robot parameters according to the platform type corresponding to the node; Step 3: The modular edge computing control device (100) performs a self-test after being powered on; Step 4: After the self-test is completed, the modular edge computing control device (100) registers the node with the global control master station (200) through the first communication link module (150); Step 5: After registration is completed, each node sends heartbeat information through the first communication link module (150) according to a preset cycle; Step 6: Transmit control commands, status information and heartbeat information through the first communication link module (150), and transmit image data or video data through the second communication link module (160); Step 7: When the UPS power management board (140) detects an abnormality in the main power supply, a power replacement, or a voltage fluctuation, the power path switching unit (144) switches the output terminal (145) to the backup power supply. The general edge computing baseboard (110) continues to operate and reports the abnormal power supply status to the global control master station (200) through the first communication link module (150). Step 8: When the second communication link module (160) malfunctions, the general edge computing backplane (110) stops, reduces the bit rate, or buffers image or video data, and continues to maintain the transmission of control commands, status information, and heartbeat information through the first communication link module (150); Step 9: The global control master station (200) performs tasks such as task assignment, task pause, task resumption, node switching, or collaborative path adjustment based on the platform registration information, heartbeat information, power status, link status, and task status of each node.

8. The networking method according to claim 7, characterized in that, The self-test described in step 3 includes: Step 31: The general edge computing baseboard (110) detects the status of the processor, storage, camera interface, first communication link module (150), and second communication link module (160); Step 32: The flight controller / motion control expansion board (120) detects the status of the microcontroller (121), actuator interface (122), sensor interface and power detection interface (123); Step 33: The UPS power management board (140) detects the status of the main power supply, backup power supply and output terminal (145); Step 34: The external sensor board (130) outputs the sensor's online status or calibration status.

9. The networking method according to claim 7, characterized in that, The registration information mentioned in step 4 includes at least one of the following: node identity, platform type, communication address, software version, sensor status, power status, and available link status.

10. The networking method according to claim 7, characterized in that, The heartbeat information mentioned in step 5 includes at least one of the following: node online status, current task status, power status, location or motion status summary, and link status.

Images