Integrated sensor integrated control device of robot with body

Abstract

The invention relates to an integrated sensor integrated control device of a robot with a body. According to the application, a plurality of sensors used for collecting multi-modal sensor data are integrated on the function daughter board module, and the function daughter board module and the carrier board layer are detachably connected through the board-to-board connector contained in the standardized connection interface array of the carrier board layer. The carrier plate layer further comprises a replaceable computing power chip used for carrying out fusion calculation and resource scheduling on sensor data. The board-to-board connector supports replacement of the replaceable computing power chip, and the computing power chip used for data processing can be replaced according to scene requirements without modifying a carrier board circuit. According to the invention, scattered sensors are integrated into a modularized and standardized hardware system, so that the integration complexity is reduced. The computing power chips are flexibly switched according to scene requirements, the situation of computing power redundancy or insufficient adaptation is reduced, accurate matching of computing power resources can be achieved, and the adaptability of the robot to different scenes is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence control, in particular to an integrated sensor integrated control device of a body robot. BACKGROUND

[0002] In order to realize the core functions such as environment perception, motion control, positioning and navigation, the body robot usually needs to integrate multiple types of sensors. However, the sensor system of the existing body robot generally adopts a decentralized deployment architecture, and various types of sensors exist in the form of independent modules, each sensor is configured with a dedicated control circuit, a power supply unit and a data interface, etc. In order to realize the collaborative work of multiple types of sensors, multiple mounting spaces need to be reserved on the body of the body robot, which not only leads to the increase of the overall volume and weight of the body robot, and it is difficult to meet the size requirements of lightweight robots, the decentralized hardware modules need to be connected to the control unit by multiple independent cables, which is messy and has a large number, which is easy to cause cable loosening, signal interference and hardware failure. And the computing power chip of the traditional body robot is only suitable for fixed type of sensor subboard, which is easy to cause insufficient adaptation or waste of redundancy, and is difficult to adapt to the differentiated needs of different scenes of the body robot. SUMMARY

[0003] The purpose of the present application is to provide an integrated sensor integrated control device of a body robot, which solves the problem of dispersed sensor hardware and limited adaptation scene of the traditional body robot.

[0004] In order to achieve the above purpose, the present application provides an integrated sensor integrated control device, which comprises a functional subboard module and a carrier plate layer which can be detachably connected; The functional subboard module comprises at least one functional subboard, and the functional subboard is integrated with multiple sensors for collecting multi-modal sensor data; The carrier plate layer comprises: Replaceable computing power chip for fusion computing and resource scheduling of the sensor data; Standardized connection interface array, including a plurality of board-to-board connectors supporting multiple interface types, the board-to-board connectors support the replacement of the replaceable computing power chip and the replacement and expansion of the functional subboard module.

[0005] The beneficial effects of the present application are: The application integrates multiple sensors for collecting multi-modal sensor data into a functional sub-board module, and establishes detachable connection between the functional sub-board module and a carrier board layer through a plate-to-plate connector contained in a standardized connection interface array of the carrier board layer. The carrier board layer also includes a replaceable computing power chip for fusion computing and resource scheduling of the sensor data. The plate-to-plate connector supports replacement of the replaceable computing power chip, and the computing power chip for data processing can be replaced according to the scene requirements without modifying the carrier board circuit.

[0006] The application breaks the mode of scattered deployment and independent control of multi-modal sensors in traditional embodied robots by means of the integrated sensor integrated control device composed of the detachably connected functional sub-board module and carrier board layer, integrates the scattered sensors into a modular and standardized hardware system, saves the complex wiring installation steps, and reduces the integration complexity. According to the scene requirements, the computing power chip can be flexibly switched to reduce the situation of computing power redundancy or insufficient adaptation. Compared with the scheme in which the computing power chip and the sensor are bound and need to be replaced as a whole, the replaceable computing power chip can realize accurate matching of computing power resources, and improve the adaptability of the embodied robot to different scenes.

[0007] Other features and advantages of the application will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 FIG. 1 is a structural schematic diagram of an integrated sensor integrated control device provided in an embodiment of the application; Figure 2 FIG. 2 is a structural schematic diagram of an integrated sensor integrated control device provided in another embodiment of the application.

[0009] REFERENCE SIGNS 1, functional sub-board module; 11, visual sensor sub-board; 12, multi-source sensor processing sub-board; 121, inertial measurement instrument; 122, environmental sensor; 123, satellite positioning sensor; 124, mobile network subunit; 125, wireless and Bluetooth subunit; 126, on-board Ethernet switch chip; 2, carrier board layer; 21, replaceable computing power chip; 211, hardware encoding and decoding unit; 22, standardized connection interface array; 221, plate-to-plate connector; 222, body recognition unit; 23, expansion interface; 24, data bus controller; 25, power management unit. DETAILED DESCRIPTION

[0010] 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.

[0011] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified. In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to implement and use this application. In the following description, details are set forth for illustrative purposes. It should be understood that those skilled in the art will recognize that this application can be implemented without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid unnecessary detail that would obscure the description of this application. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0012] Figure 1 This application provides an integrated sensor control device for a unibody robot in one embodiment. For example... Figure 1 As shown, the integrated sensor control device may include a detachably connected functional sub-board module 1 and a carrier board layer 2. This embodiment achieves mechanical fixation and electrical connectivity through a pluggable detachable connection, fulfilling the functional requirements of integrated multimodal sensing, flexible computing power adaptation, and rapid deployment across scenarios. The overall size of the integrated sensor control device in this embodiment can be 11×7.6cm, and can be scaled down or expanded according to the scenario. Thus, it occupies only a minimal installation space within the embodied robot body, adapting to various embodied robot forms from small drones to industrial robotic arms, overcoming the limitations of traditional sensor integration which is large in size and heavy in weight.

[0013] The functional subboard module 1 is a core execution layer for multi-modal sensor data collection and communication interaction. The functional subboard module 1 includes at least one functional subboard, and the functional subboard is integrated with multiple sensors for collecting multi-modal sensor data. The functional subboard module 1 integrates multiple types of sensors separately deployed in the traditional scheme on a standardized functional subboard, breaking the pain points of scattered deployment, independent wiring, and poor compatibility of traditional embodied robots. As an example, the functional subboard module 1 can synchronize the timestamps of the collected multi-modal sensor data and perform preliminary signal processing. For example, the visual data of the visual sensor can be denoised, and the attitude data of the inertial measurement instrument can be filtered, to reduce the processing pressure of the carrier board layer 2 later.

[0014] The carrier board layer 2 is the control center and computing core of the integrated sensor integrated control device, thereby solving the problem of scattered computing power of traditional embodied robots. The carrier board layer 2 can include a replaceable computing chip 21 and a standardized connection interface array 22. Transmitting multi-modal sensor data to the carrier board layer 2 for unified data processing can realize multi-data parallel processing and dynamic computing power scheduling, without frequent interaction through the software layer, reducing the delay of data fusion, and improving the real-time decision-making of embodied robots in applications.

[0015] The replaceable computing chip 21 is the computing core of the carrier board layer 2, which can be replaced with different computing chips according to scene requirements. The replaceable computing chip 21 is used for sensor data fusion calculation and resource scheduling. For example, the data of the visual sensor and the attitude data of the inertial measurement instrument can be fused. For another example, visual processing computing power can be preferentially allocated for high-speed obstacle avoidance tasks.

[0016] In one example, it is assumed that the current black sesame A1000 chip with a computing power of 56 TOPS (Tera Operations Per Second) that meets the multi-modal data fusion requirements of the embodied robot is used, and different computing chips such as Jetson Orin, Qualcomm RB6, and Horizon J6 can be replaced according to scene requirements. For example, a 28 TOPS lightweight version can be replaced in a lightweight scene, which reduces the overall power consumption of the device while ensuring basic perception functions. For another example, a 112 TOPS high-performance version can be replaced in a high-performance scene, which can meet the real-time processing of multi-channel video and high computing power of complex Artificial Intelligence (AI) model operations.

[0017] The standardized connection interface array 22 can include a board-to-board connector 221 supporting multiple interface types, which supports replacement of the replaceable computing power chip 21 and replacement and expansion of the functional subboard module 1. The standardized connection interface array 22 is a standardized bridge connecting the carrier board layer 2 and the functional subboard module 1, is a collection of physical interfaces, protocol specifications and supporting circuits integrated in the carrier board layer 2 and having a unified interface standard, and is used to provide a standardized docking basis for replacement of the replaceable computing power chip 21 and replacement and expansion of the functional subboard module 1. Through standardized design, different models of computing power chips and different functional subboards can be compatible and docked with the carrier board layer 2 without modifying the carrier board basic circuit, greatly reducing the cost and period of hardware upgrade. Moreover, the standardized connection interface array 22 supports full-scene interaction between the carrier board layer 2 and external modules, such as a high-speed data transmission interface, a control signal interface, a power supply interface, etc., and the pin definitions and transmission protocols of all interface types follow a unified standard.

[0018] The board-to-board connector 221 as the core physical execution component of the standardized connection interface array 22 can realize plug-in physical connection of the replaceable computing power chip 21 of the carrier board layer 2 and can also realize plug-in physical connection of the carrier board layer 2 and the functional subboard module 1. The board-to-board connector 221 has a standardized pin arrangement and mechanical structure, and the number, pitch and interface form of the pins match the requirements of computing power chip replacement and functional subboard docking. Through the pins of the board-to-board connector 221, the circuit signal of the carrier board layer 2 can be transmitted to an external module, and sensor data of the external module can also be transmitted back to the carrier board layer 2 through the board-to-board connector 221.

[0019] Since the pin definitions of the board-to-board connector 221 are unified, the pin positions of different computing power chips are completely matched, and replacement can be completed only by plugging. Correspondingly, the functional subboard module 1 can include edge pins, a row of metal conductive pins distributed on the edge of the functional subboard module, which are designed to match the board-to-board connector 221 and are a physical medium for mechanical fixation and electrical communication between the functional subboard module 1 and the carrier board layer 2. Through plug-in docking of the edge pins and the board-to-board connector 221, power transmission, data interaction and control signal transmission between the functional subboard module 1 and the carrier board layer 2 can be completed. Through quick plugging, when the computing power chip or the functional subboard needs to be replaced, the carrier board circuit can be disassembled only by separating the board-to-board connector 221 without welding or damaging the carrier board circuit, realizing modular assembly and facilitating subsequent maintenance.

[0020] Due to the magnitude difference in data volume, transmission rate, protocol requirement, power supply requirement, and extension maintenance requirement of the visual sensor and other sensors, in order to reduce the occupation of high-speed data to low-speed channel and the situation that low-speed data slows down high-speed transmission, the embodiments of the present application can integrate the visual sensor and other sensors on different sub-boards. Under the framework of integration, the characteristics of different sensors are designed, which can not only retain the advantages of multi-sensor centralized control, but also solve the pain points of technical conflicts of different types of sensors, and realize efficient, flexible and reliable integration mode.

[0021] Figure 2 FIG. 1 is a structural schematic diagram of an integrated sensor integration control device provided in another embodiment of the present application. As shown in FIG. 1, the functional sub-board module 1 can include a visual sensor sub-board 11 and a multi-source sensor processing sub-board 12. The visual sensor sub-board 11 can use a high-speed bus alone to transmit high-bandwidth video streams. The multi-source sensor processing sub-board 12 can use a low-speed bus to transmit low-bandwidth data. Figure 2

[0022] The visual sensor sub-board 11 can be connected to the board connector 221 through the differential signal pin connector. The visual sensor sub-board 11 can include an internal visual interface and an external visual interface connected to the visual sensor. The visual sensor connected to the visual sensor sub-board through the external visual interface is extended to the shell mounting position of the embodied robot through a shielded cable.

[0023] The visual sensor sub-board 11 is used to collect high-bandwidth visual data, and provides visual data for the embodied robot to obtain environmental visual information (such as obstacle recognition and scene positioning, etc.). The differential signal pin is an anti-interference carrier for high-speed data transmission. Through a pair of reverse transmission signals, external electromagnetic interference can be cancelled. Compared with single-ended signals, the anti-interference ability of differential signals is higher, and the signal transmission loss is lower, which can reduce the frame loss and error code conditions of visual data transmission between the functional sub-board module 1 and the carrier board layer 2. The dual design of the internal visual interface and the external visual interface can cover the full scene requirement of the embodied robot sensor. The internal visual interface can be used for visual sensors installed close to the functional sub-board module 1, and can be connected to the visual sensor through the on-board flexible cable, which is suitable for small robots. The external visual interface can be used for visual sensors that need to be extended to the shell mounting position of the embodied robot, and can be connected through a shielded cable. In this way, the long-distance and dispersed installation requirements of visual sensors of large industrial embodied robots can be met.

[0024] ​The multi-source sensor processing subboard 12 is used for integrating multiple sensors and multiple communication processing units to the same circuit board, and the multiple sensors and the multiple communication processing units are connected to the board-to-board connector 221 according to the interface type. The multi-source sensor processing subboard 12 is an integrated subboard of non-visual multi-modal sensors and communication interaction, and can integrate the traditional scattered non-visual sensors (such as inertial measurement instruments, environmental sensors, satellite positioning sensors, etc.) and the communication processing units (such as mobile network subunits, wireless and Bluetooth subunits, and on-board Ethernet switch chips, etc.) to a single circuit board, and through interface adaptation and the board-to-board connector 221, the problems of scattered sensors, complex wiring and poor communication compatibility of traditional embodied robots can be solved, and the motion control and remote interaction of the embodied robot can be realized.

[0025] In one example, the multiple sensors integrated by the multi-source sensor processing subboard 12 can include an inertial measurement instrument 121, an environmental sensor 122, and a satellite positioning sensor 123. The multiple communication processing units integrated by the multi-source sensor processing subboard 12 can include a mobile network subunit 124, a wireless and Bluetooth subunit 125, and an on-board Ethernet switch chip 126.

[0026] Due to the extremely small amount of sensor data and the differentiated real-time requirements, the inertial measurement instrument 121 and the environmental sensor 122 can be connected to the board-to-board connector 221 through an I2C bus interface, for real-time transmission of motion posture data of the embodied robot, and the satellite positioning sensor 123 can be connected to the board-to-board connector 221 through a universal asynchronous receiver / transmitter (UART) interface, for outputting positioning data of the embodied robot. In the traditional scheme, the inertial measurement instrument 121 and the environmental sensor 122 need to occupy 8-12 pins, which easily leads to messy wiring, and the parallel of multiple sensors easily leads to signal crosstalk. In the embodiment of the present application, the I2C bus needs only two pins, so that the multi-device cascade of the inertial measurement instrument 121 and the environmental sensor 122 can be realized, the pin resources are saved, the wiring density is reduced, and the risk of electromagnetic interference is also reduced. In this way, while saving interface resources, the accurate transmission of motion posture and positioning data can be ensured. In the traditional scheme, the satellite positioning sensor synchronous interface needs strict clock matching, but the vibration of the embodied robot easily leads to clock offset and data packet loss. In the embodiment of the present application, the UART does not need a clock line, and the data transmission can be realized through the frame structure of the start bit, the data bit and the stop bit, which can improve the anti-vibration interference capability, reduce the packet loss rate, meet the continuous positioning requirement of the embodied robot in the outdoor environment without interruption, and can support long-distance transmission.

[0027] In view of the large amount of data of the communication processing unit and the differentiated rate requirements, the mobile network subunit 124 can be connected to the board-to-board connector 221 through the PCIe interface of the multi-source sensor processing subboard 12. The wireless and Bluetooth subunit 125 can be connected to the board-to-board connector 221 through the Secure Digital Input and Output (SDIO) interface. The on-board Ethernet switch chip 126 can be connected to the board-to-board connector 221 through the Reduced Gigabit Media Independent Interface (RGMII) interface. The traditional mobile network bandwidth is small, and it is difficult to meet the large flow requirements of multi-channel video backhaul and remote control instruction interaction. Through the PCIe interface, high bandwidth can be ensured, and low delay interaction and remote control reliability can be ensured. If the wireless and Bluetooth transmission adopts the PCIe interface, it is easy to cause waste of interface resources, therefore, through the SDIO interface, the medium-speed transmission requirements can be adapted, and the complexity of hardware design is reduced. The transmission of Ethernet is usually complex in the traditional scheme, and has no switch function, and can only connect one external device, which is difficult to meet the multi-device cooperation of industrial robots. Therefore, through the RGMII interface, the data timing compression is reduced, the subboard wiring difficulty is reduced, and full-duplex transmission is supported, meeting the multi-device parallel data interaction. Through the connection with the carrier board layer 2, a plurality of external Ethernet ports can be expanded, without the need for additional deployment of a switch, reducing the integration cost. At the same time, the transmission delay of RGMII is small, and the response of the control instruction can be timely. In this way, multi-link parallel transmission of mobile network, wireless and Ethernet can be realized.

[0028] In the embodiment of the present application, the functional subboard module 1 can further include a calibration unit, which is electrically connected to a plurality of sensors of the multi-source sensor processing subboard 12 through a hardware calibration circuit, for compensating the measurement error of the sensors in real time through the on-board dedicated hardware logic, eliminating the error from the hardware level, without relying on the post-processing of the software in the carrier board layer 2, and providing high-precision data for the fusion of the multi-modal data of the carrier board layer 2.

[0029] As an example, the hardware calibration circuit of the calibration unit can adopt a design of one type of sensor corresponding to one set of special logic, and customize the compensation scheme according to the error characteristics of different sensors. The hardware calibration circuit can include a drift error calibration circuit corresponding to the inertial measurement instrument 121 and the environmental sensor 122, an image distortion calibration circuit corresponding to the visual sensor, and an altitude measurement cross verification calibration circuit corresponding to the satellite positioning sensor 123.

[0030] The measurement error of the inertial measurement instrument 121 mainly comes from temperature drift and time drift, and the measurement error of the environment sensor 122 mainly comes from temperature drift. Therefore, the embodiment of the application can integrate a high-precision temperature sensor and a special calibration ASIC chip, and embed a solidified logic of error compensation algorithm and a nonvolatile storage medium. The temperature sensor collects the working temperature of the inertial measurement instrument 121 and the environment sensor 122 in real time, for example, updated once every 1 ms, and the calibration ASIC chip calls the temperature-error mapping table stored in the nonvolatile storage medium according to the current working temperature, calculates a real-time compensation coefficient, and then directly superimposes the compensation coefficient into the original data of the sensor through hardware logic without software processing.

[0031] The image distortion of the visual sensor mainly includes radial distortion caused by the edge pixel deviation of lens curvature and tangential distortion caused by the pixel tilt due to lens installation deviation. The traditional software calibration needs to correct the pixels of each frame of image through software, which occupies a lot of computing power. The embodiment of the application can calibrate the image distortion through the image distortion calibration circuit. The distortion calibration circuit can include a field programmable gate array (FPGA), a storage unit of distortion parameters and an image cache unit. For example, the pixels can be corrected in real time through the FPGA. The original image output by the visual sensor can be first transmitted to the image cache unit, the FPGA calls the coefficients in the storage unit of distortion parameters, calculates the correction coordinates of each pixel in real time through hardware logic, and then directly outputs the corrected image data to the board layer 2. The whole process is completed in the hardware layer and does not occupy software computing power.

[0032] The altitude measurement error of the satellite positioning sensor 123 is large and is prone to data jump caused by shielding. The altitude measurement cross-verification calibration circuit of the embodiment of the application can include a double-port data comparator, a Kalman filter hardware acceleration unit and a confidence evaluation circuit. The double-port data comparator receives the altitude data received by the satellite positioning sensor 123 and the altitude data calculated by the barometer, and then the confidence evaluation circuit judges the data validity of the satellite positioning sensor 123. Then, the Kalman filter hardware acceleration unit dynamically allocates weights according to the confidence to output the calibrated altitude in real time. The fusion process is completed through hardware logic.

[0033] The calibration unit can eliminate the error of the sensor from the hardware layer, and improve the accuracy of the sensor data. Moreover, the hardware real-time calibration does not need to rely on the software algorithm of the carrier board layer 2, which improves the calibration efficiency and releases the computing power resources of the carrier board layer 2. The hardware calibration logic solidification can also ensure the stability and consistency of the calibration, and the calibration unit is integrated in the functional subboard module 1, without the need for additional cable arrangement, without increasing the size of the device, and the calibration parameters can be stored through the non-volatile storage medium. The functional subboard does not need to be re-calibrated when it is replaced, and the plug-and-play effect can be achieved.

[0034] In the embodiments of the present application, the carrier board layer 2 can also include a unified expansion interface 23, which is a standardized module interface with multiple pins, and the pins of the expansion interface have a bidirectional connection relationship. As an example, the expansion interface can be a unified expansion interface with 40 pins. The pin pitch can be set to an industrial precision pitch to ensure the stability of the connection, for example, 0.8 mm. The pin material can be gold-plated copper to reduce the risk of oxidation and improve the conductivity and service life.

[0035] In the embodiments of the present application, the expansion interface 23 can be connected to the body of the embodied robot and / or external equipment through a cable, and is the only interface to the outside, which is used to simplify all communication and data exchange between the robot body and the robot body. The expansion interface 23 can adapt to different types of embodied robot bodies and / or external equipment through the type of multiple pins. As an example, the expansion interface 23 can include at least GPIO pins for transmitting on-off control signals, Controller Area Network (CAN) bus pins for industrial real-time control data transmission, industrial serial bus (such as RS485) pins for long-distance low-speed data transmission, power supply pins, ground pins, and spare expansion pins.

[0036] In the embodiments of the present application, the carrier board layer 2 can also include a data bus controller 24. The data bus controller 24 is integrated with the physical transmission lines of the data bus through the on-board wiring of the carrier board layer 2 (such as unified management of high-speed differential signal lines and low-speed control signal lines), and extends the transmission channel of the data bus to the functional subboard module 1 through the board-to-board connector 221. The data bus controller 24 integrates the control logic and the physical link, and through the unified bus control logic and the extended high-speed transmission link, the data interaction between the carrier board layer 2 and the functional subboard module 1 can be realized. multiple types, high bandwidth and low delay.

[0037] For example, an industrial-grade PCIe switch chip is adopted to support multiple PCIe channels, and control logic compatible with I2C, UART, CAN and other low-speed buses is adopted to realize unified scheduling of high-speed data and low-speed control. A high-precision crystal oscillator and a clock buffer are integrated to provide a synchronous clock signal for the bus transmission link, ensuring the consistency of the transmission time of multiple subboard data. Through a signal conditioning circuit containing a signal amplifier and an equalizer, the signal attenuation of the onboard wiring and the board-to-board connector 221 is compensated. Moreover, by monitoring the bus transmission state in real time, monitoring data can be fed back to the replaceable computing chip 21 to dynamically adjust the transmission strategy.

[0038] Through the high-speed bus and dynamic bandwidth allocation, the embodiment of the present application can simultaneously support the parallel transmission of multiple types of data, meeting the real-time requirements of the body robot for high-speed obstacle avoidance and dynamic positioning. The high-speed, low-speed and industrial control buses are uniformly managed, and the carrier board layer 2 only needs one set of bus to cover all the needs of the subboards, reducing the density of wiring and electromagnetic interference. At the same time, the functional subboard module 1 does not need to design multiple sets of bus interfaces, and through the pins of the unified board-to-board connector 221, the complexity of hardware design can be reduced, and the expansion of functional subboards can also be performed. Through the unified control logic, the problem of scattered bus congestion and synchronization difference can be solved, providing bottom support for high-precision data fusion of the carrier board layer and control of the body robot.

[0039] In the embodiment of the present application, the carrier board layer 2 can also include a power management unit 25 configured with multiple adjustable power supply channels. The power supply channels have a mapping relationship with the multiple source circuits in the functional subboard module 1. For example, a one-to-one or multiple-to-one mapping relationship is formed with the multiple source circuits such as the visual sensor subboard 11, the multiple source sensor processing subboard 12 and the expansion interface 23 of the functional subboard module 1. The power management unit 25 is used to control the power output and on-off state of the power supply channel according to the task type of the body robot, to control the distribution of the multiple source circuits in the functional subboard module 1, and to realize the on-demand distribution and precise control of power resources. In this way, both the excessive consumption of electrical energy can be reduced, and the stability of power supply in extreme scenarios can be ensured.

[0040] In the embodiment of the present application, the replaceable computing chip 21 can include a hardware encoding and decoding unit 211 for performing encoding and corresponding decoding operations on sensor data through built-in hardware circuits.

[0041] The hardware encoding and decoding unit 211 performs real-time encoding and decoding on the original sensor data through built-in hardware circuits. After the sensors of the functional subboard module 1 generate the original sensor data, the data can be transmitted to the carrier board layer 2 through the board-to-board connector 221. The replaceable computing chip 21 of the carrier board layer 2 performs encoding and / or decoding operations through the hardware encoding and decoding unit 211.

[0042] For example, for the high efficiency video coding standard H.265 or the advanced video coding standard H.264 hardware codec, a new generation of video coding standard such as AV (AOMedia Video) 1, VP (Video Processing) 9 or High Efficiency Video Coding (HEVC) can be used. The integration of different computing power chips and the principle of hardware acceleration are the same.

[0043] In the embodiment of the present application, the replaceable computing power chip 21 can also be used to identify the environment information of the embodied robot in real time, and to judge the environment in which the embodied robot is located in real time. Then, the fusion weight of the sensor data is dynamically adjusted based on the environment information. For example, the positioning data signal of the indoor closed space is weak, and the fusion weight of the visual data and the attitude data can be increased; the positioning data signal of the outdoor open area is strong, and the weight of the positioning data can be increased; the vibration of the strong interference industrial environment is large, and the weight of the attitude data can be reduced, and the weight of the visual data can be increased. In addition, the replaceable computing power chip 21 can generate a fusion algorithm adapted to the current environment according to the adjusted fusion weight, and transmit the fusion algorithm corresponding to the sensor data after adjusting the fusion weight to the hardware codec unit 211. The hardware codec unit 211 can also be used to solidify the fusion algorithm through the built-in hardware circuit. In this way, without software instructions, the fusion algorithm process can be directly realized through the circuit logic, and the solidification of the fusion algorithm can reduce the delay of the calculation. The replaceable computing power chip 21 continuously monitors the change of the environment, and if the scene switching is detected, the hardware level calculation can be repeated, and the dynamic update of the fusion algorithm can be realized.

[0044] In the embodiment of the present application, the standardized connection interface array 22 can also include a body recognition unit 222, which communicates with the body of the embodied robot through the on-board hardware signal interaction circuit. The body recognition unit 222 establishes electrical connection with the body of the embodied robot through the on-board hardware signal interaction circuit, such as a special recognition pin and an encrypted communication module, to complete the embodied robot type recognition and control parameter configuration process, so as to realize plug and play with different types of embodied robot bodies.

[0045] The body recognition unit 222 first acquires the hardware identification signal of the body of the accessed embodied robot, and identifies the type of the embodied robot based on the hardware identification signal. The body recognition unit 222 can be automatically activated when the standardized connection interface array 22 is detected to be connected with the body of the embodied robot. The body recognition unit can identify the hardware identification signal of the embodied robot through, for example, resistance coding recognition, pin level combination, and encrypted chip interaction, and determine the type of the robot based on the hardware identification signal.

[0046] Then, based on a mapping library set in advance, the control parameters corresponding to the type of the embodied robot are identified. The mapping library is a mapping library constructed based on the mapping relationship between the type of the robot and the control parameters. The control parameters can include sensor parameters, communication protocols, and computing power allocation strategies. Finally, based on the control parameters, configuration instructions are sent to the replaceable computing power chip 21 to initialize the control parameters of the embodied robot. For example, after receiving the configuration instructions, the replaceable computing power chip 21 can send sensor calibration parameters to the multi-source sensor processing sub-board 12, configure the communication protocol of the data bus controller 24, adjust the algorithm parameters of the hardware encoding and decoding unit 211, and allocate computing power resources, etc.

[0047] Through the body recognition unit 222, the automatic recognition and parameter configuration of the embodied robot can be completed, reducing the error rate of manual configuration and improving the service life of the equipment. Based on the pre-stored standardized setting parameters, the initial parameters of the embodied robots of the same type can be consistent, improving the stability of work. In addition, it can also reduce the situation of data leakage or equipment misoperation caused by the access of unauthorized embodied robots, and improve the security of information.

[0048] The integrated sensor integrated control device in the embodiments of the present application can be used in many subdivided fields and scenarios. For example, in an automatic driving auxiliary system, it can be applied to intelligent driving assistance of passenger cars and commercial vehicles, realizes lane keeping, automatic parking and other functions by using integrated visual sensors and inertial measuring instruments, and realizes real-time environment perception by hardware codec processing. In a UAV flight control system, it can be applied to flight control of consumer and industrial UAVs, realizes precise hovering and autonomous flight by combining satellite positioning sensors, inertial measuring instruments and visual sensors, and realizes low-delay attitude control by using hardware acceleration. In intelligent security monitoring, it can be applied to intelligent monitoring systems of parks and buildings. It realizes all-around monitoring by using multiple visual sensors, and realizes efficient video encoding and transmission by hardware codec. In industrial automation control, it can be applied to automation control of production lines and warehouse systems. It realizes precise positioning and environment perception by using a sensor integrated module, and is seamlessly integrated with industrial equipment through a CAN bus. In a blind person navigation auxiliary system, it can be applied to daily travel assistance of visually impaired people. It provides safe navigation for the blind by combining voice prompts and tactile feedback with existing map data, and is particularly suitable for barrier-free path planning in complex indoor and outdoor environments. The common point of all application scenarios is that the core technical architecture of integrated sensor integration and hardware codec processing remains unchanged, and only the specific sensor configuration, computing power specification and application scenario are different.

[0049] Those skilled in the art can understand that all or part of the functions of the various methods in the above embodiments can be realized by hardware or by a computer program. When all or part of the functions in the above embodiments are realized by a computer program, the program can be stored in a computer readable storage medium, which can include read-only memory, random access memory, magnetic disk, optical disk, hard disk, etc. The above functions are realized by executing the program by a computer. For example, the program is stored in the memory of the device, and when the program in the memory is executed by the processor, the above functions are realized. In addition, when all or part of the functions in the above embodiments are realized by a computer program, the program can also be stored in a server, another computer, a storage medium such as a disk, an optical disk, a flash disk or a mobile hard disk, and is downloaded or copied into the memory of the local device, or the system of the local device is updated, and when the program in the memory is executed by the processor, the above functions are realized.

[0050] The above application uses specific examples to illustrate the present application, which is only used to help understand the present application and does not limit the present application. For those skilled in the art to which the present application belongs, according to the idea of the present application, some simple deductions, deformations or substitutions can be made.

Claims

1. An integrated sensor control device for a unibody robot, characterized in that, Includes detachable functional sub-board modules and carrier board layers; The functional subboard module includes at least one functional subboard, which integrates multiple sensors for collecting multimodal sensor data. The carrier plate layer includes: A replaceable computing chip is used for fusion computing and resource scheduling of the sensor data; A standardized connection interface array includes board-to-board connectors that support multiple interface types. These board-to-board connectors support the replacement of the replaceable computing chip and the replacement and expansion of the functional sub-board modules.

2. The integrated sensor control device according to claim 1, characterized in that, The functional sub-board module includes: The vision sensor sub-board is connected to the board-to-board connector via differential signal pins. The vision sensor sub-board includes an internal vision interface and an external vision interface for connecting the vision sensor. The vision sensor connected to the vision sensor sub-board via the external vision interface extends to the outer shell mounting position of the embodied robot via a shielded cable. A multi-source sensor processing sub-board is used to integrate multiple sensors and multiple communication processing units onto the same circuit board, wherein the multiple sensors and multiple communication processing units are connected to the board-to-board connector according to the adapted interface type.

3. The integrated sensor control device according to claim 2, characterized in that, The multi-source sensor processing subboard integrates various sensors, including: An inertial measurement unit and environmental sensors are connected to the board-to-board connector via an I2C bus interface for real-time transmission of the robot's motion posture data. A satellite positioning sensor is connected to the board-to-board connector via a UART interface to output positioning data of the embodied robot. The multi-source sensor processing subboard integrates various communication processing units, including: The mobile network subunit is connected to the board-to-board connector via the PCIe interface of the multi-source sensor processing subboard; The wireless and Bluetooth sub-units are connected to the board-to-board connector via an SDIO interface; An onboard Ethernet switch chip is connected to the board-to-board connector via an RGMII interface.

4. The integrated sensor control device according to claim 3, characterized in that, The functional sub-board module also includes a calibration unit, which is electrically connected to various sensors on the multi-source sensor processing sub-board through a hardware calibration circuit, and is used to compensate for the measurement error of the sensors in real time through onboard dedicated hardware logic; The hardware calibration circuit includes a drift error calibration circuit for the inertial measurement instrument and environmental sensor, an image distortion calibration circuit for the visual sensor, and an altitude measurement cross-validation calibration circuit for the satellite positioning sensor.

5. The integrated sensor control device according to claim 1, characterized in that, The carrier board layer also includes a unified expansion interface, which is a standardized multi-pin module interface, and the pins of the expansion interface have a bidirectional connection relationship. The expansion interface connects to the body of the embodied robot and / or external devices via a cable, and the expansion interface is compatible with different types of the body of the embodied robot and / or external devices. The expansion interface includes at least GPIO pins for transmitting switch control signals, CAN bus pins for industrial-grade real-time control data transmission, RS485 serial port pins for long-distance low-speed data transmission, power supply pins, ground pins, and spare expansion pins.

6. The integrated sensor control device according to claim 1, characterized in that, The replaceable computing chip includes: The hardware encoding / decoding unit is used to encode and decode the sensor data using built-in hardware circuitry.

7. The integrated sensor control device according to claim 6, characterized in that, The replaceable computing chip is also used to identify the environmental information of the embodied robot in real time, dynamically adjust the fusion weight of the sensor data based on the environmental information, and transmit the fusion algorithm corresponding to the sensor data after adjusting the fusion weight to the hardware encoding and decoding unit; the hardware encoding and decoding unit is also used to solidify the fusion algorithm through the built-in hardware circuit.

8. The integrated sensor control device according to claim 1, characterized in that, The carrier plate layer also includes: The data bus controller is integrated with the physical transmission line of the data bus through the onboard wiring of the carrier board layer, and extends the transmission channel of the data bus to the functional sub-board module through the board-to-board connector.

9. The integrated sensor control device according to claim 1, characterized in that, The carrier board layer also includes a power management unit, which is configured with multiple adjustable power supply channels. The power supply channels are mapped to the multi-source circuits in the functional sub-board module. The power management unit is used to control the power output and on / off state of the power supply channels according to the task type of the embodied robot, so as to control the allocation of the multi-source circuits in the functional sub-board module.

10. The integrated sensor control device according to claim 1, characterized in that, The standardized connection interface array includes a body recognition unit, which communicates with the body of the embodied robot through an onboard hardware signal interaction circuit. The entity identification unit is used for: Obtain the hardware identification signal of the body of the connected embodied robot, and identify the type of the embodied robot based on the hardware identification signal; Based on a pre-set mapping library, control parameters corresponding to the type of the embodied robot are identified. These control parameters include sensor parameters, communication protocols, and computing power allocation strategies. Based on the control parameters, a configuration command is sent to the replaceable computing chip to initialize the control parameters of the embodied robot.

Images