A mission-driven autonomous unmanned equipment prototype system

Abstract

The present invention belongs to the field of computer and embedded technology related to robots, and specifically relates to a task-driven autonomous unmanned equipment prototype system. The task-driven autonomous unmanned equipment prototype system includes: a CPU computing processing unit, an FPGA interface management unit, and a system evolution management unit; the present invention proposes a design scheme for an unmanned equipment prototype system under task drive, which meets the design requirements of various unmanned platform computing and processing systems through a typical hardware architecture of CPU+FPGA, uses task environment information to complete the matching of the optimal architecture template, and uses Multi-Boot technology to complete the reconfiguration of interface resources. For the CPU unit, the performance and power consumption are adapted by calling the operating system kernel service. The unmanned equipment prototype system achieves the optimal adaptation of performance, algorithm and task, and improves the reconfigurability and evolvability of the system.

Description

Technical Field

[0001] The present invention belongs to the field of robot-related computers and embedded technologies, and in particular relates to a task-driven autonomous unmanned equipment prototype system. Background Art

[0002] Traditional unmanned equipment mainly uses customized computing systems, whose hardware IO resources and computing processing performance are completely fixed after leaving the factory, and only the inherent resources can be modified. Unmanned equipment faces complex and changing scenarios, and the fixed system design scheme is difficult to take into account the optimal match in terms of power consumption and performance. For new unmanned platforms, it is often necessary to redesign the computing system, which increases the cost and development cycle of system design. Summary of the invention

[0003] 1. Technical issues to be resolved

[0004] The technical problem to be solved by the present invention is to solve the problems of traditional unmanned equipment system design being rigid and unable to adapt to changes in tasks and environment.

[0005] (II) Technical solution

[0006] In order to solve the above technical problems, the present invention provides a task-driven autonomous unmanned equipment prototype system, the task-driven autonomous unmanned equipment prototype system comprises: a CPU computing processing unit, an FPGA interface management unit, and a system evolution management unit;

[0007] The CPU computing unit is used to complete the unmanned platform motion control, computing, and logic processing functions;

[0008] The FPGA interface management unit is used to manage and configure input and output resources including servo interfaces and sensor interfaces;

[0009] The system evolution management unit is used to configure and reconfigure the FPGA and CPU units based on the input task and environment information.

[0010] The FPGA and the CPU unit access data from each other via a PCIE bus.

[0011] The system evolution management unit completes the firmware hot switch by operating the FPGA register.

[0012] The system evolution management unit uses the SPI bus to communicate with the CPU core to adjust the processor performance.

[0013] Wherein, the system evolution management unit includes an evolution motivation reasoning module, a system evolution reconstruction module and a system performance evaluation module;

[0014] The evolutionary motivation reasoning module is used to obtain four indicators of task difficulty, system real-time performance, autonomy, and endurance through the current environment model and task model information, and further match them to the corresponding system architecture template through fuzzy mapping;

[0015] The system evolution and reconstruction module is used to call the unit evolution tool to complete the online adjustment of FPGA resources, CPU computing power, CPU core number and main frequency;

[0016] The system performance evaluation module is used to complete the online evaluation of comprehensive computing power, power consumption, and real-time performance, and compare with the set template to verify that the prototype system completes the reconstruction evolution under task-driven.

[0017] Among them, the autonomous unmanned equipment prototype system is driven by the environment model and the task model to complete the matching of the system architecture template, thereby improving the adaptability and reconfigurability of the unmanned equipment computing system.

[0018] The system evolution steps of the autonomous unmanned equipment prototype system executed by the system evolution management unit are as follows:

[0019] Step 1: The evolutionary motivation reasoning module completes the construction of the environmental model;

[0020] Step 2: Process the environment model and task model information through the inference neural network to obtain four indicators: task difficulty, system real-time, autonomy, and endurance, and output them to the system evolution reconstruction module;

[0021] Step 3: The system evolution reconstruction module maps the input four indicator values ​​with three types of system architecture templates using fuzzy strategy logic and uses them as the evolution target of the current prototype system;

[0022] Step 4: For the FPGA interface management unit, the online hot switch of the FPGA internal firmware is completed through MultiBoot, thereby completing the reconfiguration of the IO interface resources;

[0023] Step 5: For the CPU computing processing unit, access the underlying kernel register of the operating system through SPI to complete the online configuration of the CPU power consumption mode, CPU main frequency voltage, and CPU quantity;

[0024] Step 6: Determine the prototype system architecture by reading the current interface hardware device ID and CPU processing unit ID to complete the reconstruction; investigate the system's internal monitoring software to complete the confirmation of power consumption performance and processor quantity information.

[0025] Wherein, in step 1, the evolutionary motivation reasoning module uses the SLAM algorithm to complete the construction of the environment model.

[0026] Wherein, in step 1, the environment model includes an elevation map, a global occupancy map, terrain semantic attributes, mission modes and navigation waypoints.

[0027] Among them, the autonomous unmanned equipment prototype system meets the design requirements of various unmanned platform computing and processing systems through the typical hardware architecture of CPU+FPGA, uses task environment information to complete the matching of the optimal architecture template, and uses Multi-Boot technology to complete the reconfiguration of interface resources. For the CPU unit, the performance and power consumption are adapted by calling the operating system kernel service. The unmanned equipment prototype system achieves the optimal adaptation of performance, algorithms and tasks, and improves the system's reconfiguration and evolvability.

[0028] (III) Beneficial effects

[0029] Compared with the prior art, the present invention proposes a design scheme for an unmanned equipment prototype system under task drive, meets the design requirements of various unmanned platform computing and processing systems through the typical hardware architecture of CPU+FPGA, uses task environment information to complete the matching of the optimal architecture template, and uses Multi-Boot technology to complete the reconfiguration of interface resources. For the CPU unit, the performance and power consumption are adapted by calling the operating system kernel service. The unmanned equipment prototype system achieves the optimal adaptation of performance, algorithms and tasks, and improves the system's reconfiguration and evolvability. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Schematic diagram of the mission-driven unmanned equipment prototype system.

[0031] Figure 2 Schematic diagram of the logical structure of the system evolution management unit. DETAILED DESCRIPTION

[0032] In order to make the purpose, content, and advantages of the present invention more clear, the specific implementation methods of the present invention are further described in detail below in conjunction with the accompanying drawings and examples.

[0033] In order to solve the above technical problems, the present invention provides a task-driven autonomous unmanned equipment prototype system, such as Figure 1 As shown, the task-driven autonomous unmanned equipment prototype system includes: a CPU computing processing unit, an FPGA interface management unit, and a system evolution management unit;

[0034] The CPU computing unit is used to complete the unmanned platform motion control, computing, and logic processing functions;

[0035] The FPGA interface management unit is used to manage and configure input and output resources including servo interfaces and sensor interfaces;

[0036] The system evolution management unit is used to configure and reconfigure the FPGA and CPU units based on the input task and environment information.

[0037] The FPGA and the CPU unit access data from each other via a PCIE bus.

[0038] The system evolution management unit completes the firmware hot switch by operating the FPGA register.

[0039] The system evolution management unit uses the SPI bus to communicate with the CPU core to adjust the processor performance.

[0040] Among them, Figure 2 As shown, it is a schematic diagram of the logical frame relationship structure of the system evolution management unit, and the system evolution management unit includes an evolution motivation reasoning module, a system evolution reconstruction module and a system performance evaluation module;

[0041] The evolutionary motivation reasoning module is used to obtain four indicators of task difficulty, system real-time performance, autonomy, and endurance through the current environment model and task model information, and further match them to the corresponding system architecture template through fuzzy mapping;

[0042] The system evolution and reconstruction module is used to call the unit evolution tool to complete the online adjustment of FPGA resources, CPU computing power, CPU core number and main frequency;

[0043] The system performance evaluation module is used to complete the online evaluation of comprehensive computing power, power consumption, and real-time performance, and compare with the set template to verify that the prototype system completes the reconstruction evolution under task-driven.

[0044] Among them, the autonomous unmanned equipment prototype system is driven by the environment model and the task model to complete the matching of the system architecture template, thereby improving the adaptability and reconfigurability of the unmanned equipment computing system.

[0045] The system evolution steps of the autonomous unmanned equipment prototype system executed by the system evolution management unit are as follows:

[0046] Step 1: The evolutionary motivation reasoning module completes the construction of the environmental model;

[0047] Step 2: Process the environment model and task model information through the inference neural network to obtain four indicators: task difficulty, system real-time, autonomy, and endurance, and output them to the system evolution reconstruction module;

[0048] Step 3: The system evolution reconstruction module maps the input four indicator values ​​with three types of system architecture templates using fuzzy strategy logic and uses them as the evolution target of the current prototype system;

[0049] Step 4: For the FPGA interface management unit, the online hot switch of the FPGA internal firmware is completed through MultiBoot, thereby completing the reconfiguration of the IO interface resources;

[0050] Step 5: For the CPU computing processing unit, access the underlying kernel register of the operating system through SPI to complete the online configuration of the CPU power consumption mode, CPU main frequency voltage, and CPU quantity;

[0051] Step 6: Determine the prototype system architecture by reading the current interface hardware device ID and CPU processing unit ID to complete the reconstruction; investigate the system's internal monitoring software to complete the confirmation of power consumption performance and processor quantity information.

[0052] Wherein, in step 1, the evolutionary motivation reasoning module uses the SLAM algorithm to complete the construction of the environment model.

[0053] Wherein, in step 1, the environment model includes an elevation map, a global occupancy map, terrain semantic attributes, mission modes and navigation waypoints.

[0054] Among them, the autonomous unmanned equipment prototype system meets the design requirements of various unmanned platform computing and processing systems through the typical hardware architecture of CPU+FPGA, uses task environment information to complete the matching of the optimal architecture template, and uses Multi-Boot technology to complete the reconfiguration of interface resources. For the CPU unit, the performance and power consumption are adapted by calling the operating system kernel service. The unmanned equipment prototype system achieves the optimal adaptation of performance, algorithms and tasks, and improves the system's reconfiguration and evolvability.

[0055] Example 1

[0056] This embodiment proposes a design scheme for a prototype system of unmanned equipment driven by tasks, which uses FPGA and multiple CPUs as core computing units. According to the different requirements of tasks and environments, three typical system architecture templates are established, and finally the evolution of the architecture and the adjustment of unit performance are completed through online configuration, so as to achieve the optimal adaptation to the current task scenario in terms of interface resources, computing power and power consumption.

[0057] The task-driven unmanned equipment prototype system design proposed in this embodiment uses FPGA and multiple CPUs as core computing units. According to the different requirements of tasks and environments, three typical system architecture templates are established, and the evolution of the architecture and the adjustment of unit performance are finally completed through online configuration, so as to achieve the optimal adaptation to the current task scenario in terms of interface resources, computing power and power consumption. The typical hardware system architecture design for unmanned platforms is as follows:

[0058] 1) Architecture 1, CPU processing performance is low, IO input is IMU sensor data, remote control data, joint torque and position data, IO output is servo drive instructions;

[0059] 2) Architecture 2, CPU processing performance is higher, IO input is IMU sensor data, radar depth structure data, GPS data, remote control data, joint torque and position data, IO output is servo drive instructions, load communication interface;

[0060] 3) Architecture 3, high CPU processing performance, multi-threaded processing function is required, IO input is IMU sensor data, radar depth structure data, visual data, remote control data, joint torque and position data, IO output is servo drive instructions, load communication interface;

[0061] For the above architecture templates, Architecture 1 is mainly aimed at the remote control task requirements of unmanned platforms. The robot does not have the ability to perceive the environment and make cognitive decisions, and completely adopts the results of task issuance; Architecture 2 is mainly aimed at single tasks and fixed environment scenarios. The platform has basic environmental and cognitive capabilities, and has interfaces for corresponding depth data, images and positioning data. The performance of the computing processing unit can only meet the execution of a single task; Architecture 3 is aimed at diversified tasks and complex environments, and has deep visual perception and positioning capabilities. The prototype system has multiple sensor acquisition interfaces such as radar and vision. The computing unit can support multi-threaded and multi-process processing to meet the simultaneous computing needs of multiple tasks.

[0062] After the task and environment information are input, the prototype system completes the evolution and reconstruction of the hardware architecture and units based on the system evolution management unit. The main steps are as follows:

[0063] (1) First, the evolutionary motivation module is based on environmental perception data, which mainly includes elevation maps, global occupancy maps, and terrain semantic attributes. Task data mainly includes task modes and navigation waypoints.

[0064] (2) Processing the task and environment model information through the inference neural network, obtaining four indicators: task difficulty, system real-time, autonomy, and endurance, and outputting them to the system evolution module;

[0065] (3) The system evolution module uses fuzzy strategy logic to map the four input indicator values ​​to the three types of system architecture templates and uses them as the evolution goals of the current prototype system;

[0066] (4) For the FPGA interface management unit, the online hot switch of the FPGA internal firmware is completed through MultiBoot, thereby completing the reconfiguration of the IO interface resources;

[0067] (5) For the CPU computing processing unit, access the underlying kernel registers of the operating system through SPI to complete the online configuration of the CPU power consumption mode, CPU main frequency voltage, and CPU quantity;

[0068] (6) Finally, the prototype system architecture is determined by reading the current interface hardware device ID and CPU processing unit ID to complete the reconstruction; the internal monitoring software of the survey system completes the confirmation of information such as power consumption performance and number of processors.

[0069] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A mission-driven autonomous unmanned equipment prototype system, characterized in that: The task-driven autonomous unmanned equipment prototype system includes: a CPU computing processing unit, an FPGA interface management unit, and a system evolution management unit; The CPU computing unit is used to complete the unmanned platform motion control, computing, and logic processing functions; The FPGA interface management unit is used to manage and configure input and output resources including servo interfaces and sensor interfaces; The system evolution management unit is used to configure and reconfigure the FPGA and CPU units based on the input task and environment information; Wherein, the system evolution management unit includes an evolution motivation reasoning module, a system evolution reconstruction module and a system performance evaluation module; The evolutionary motivation reasoning module is used to obtain four indicators of task difficulty, system real-time performance, autonomy, and endurance through the current environment model and task model information, and further match them to the corresponding system architecture template through fuzzy mapping; The system evolution and reconstruction module is used to call the unit evolution tool to complete the online adjustment of FPGA resources, CPU computing power, CPU core number and main frequency; The system performance evaluation module is used to complete the online evaluation of comprehensive computing power, power consumption, and real-time performance, and compare with the set template to verify that the prototype system completes the reconstruction evolution under task drive; The system evolution steps of the autonomous unmanned equipment prototype system executed by the system evolution management unit are as follows: Step 1: The evolutionary motivation reasoning module completes the construction of the environmental model; Step 2: Process the environment model and task model information through the inference neural network to obtain four indicators: task difficulty, system real-time, autonomy, and endurance, and output them to the system evolution reconstruction module; Step 3: The system evolution reconstruction module maps the input four indicator values ​​with three types of system architecture templates using fuzzy strategy logic and uses them as the evolution target of the current prototype system; Step 4: For the FPGA interface management unit, the online hot switch of the FPGA internal firmware is completed through MultiBoot, thereby completing the reconfiguration of the IO interface resources; Step 5: For the CPU computing processing unit, access the underlying kernel register of the operating system through SPI to complete the online configuration of the CPU power consumption mode, CPU main frequency voltage, and CPU quantity; Step 6: Determine the prototype system architecture by reading the current interface hardware device ID and CPU processing unit ID to complete the reconstruction; investigate the system's internal monitoring software to complete the confirmation of power consumption performance and processor quantity information.

2. The mission-driven autonomous unmanned equipment prototype system according to claim 1, characterized in that: The FPGA and the CPU unit access data with each other via the PCIE bus.

3. The mission-driven autonomous unmanned equipment prototype system as claimed in claim 2, characterized in that: The system evolution management unit completes the firmware hot switch by operating the FPGA register.

4. The mission-driven autonomous unmanned equipment prototype system as claimed in claim 3, characterized in that: The system evolution management unit uses the SPI bus to communicate with the CPU core to adjust the processor performance.

5. The mission-driven autonomous unmanned equipment prototype system as claimed in claim 4, characterized in that: The autonomous unmanned equipment prototype system is driven by the environment model and the task model to complete the matching of the system architecture template, thereby improving the adaptability and reconfigurability of the unmanned equipment computing system.

6. The mission-driven autonomous unmanned equipment prototype system as claimed in claim 5, characterized in that: In step 1, the evolutionary motivation reasoning module uses the SLAM algorithm to complete the construction of the environment model.

7. The mission-driven autonomous unmanned equipment prototype system according to claim 6, characterized in that: In step 1, the environment model includes an elevation map, a global occupancy map, terrain semantic attributes, mission modes and navigation waypoints.

8. The mission-driven autonomous unmanned equipment prototype system according to claim 6, characterized in that: The autonomous unmanned equipment prototype system meets the design requirements of various unmanned platform computing and processing systems through the typical hardware architecture of CPU+FPGA, uses task environment information to complete the matching of the optimal architecture template, uses Multi-Boot technology to complete the reconfiguration of interface resources, and completes the adaptation of performance and power consumption for the CPU unit by calling the operating system kernel service. The unmanned equipment prototype system achieves the optimal adaptation of performance, algorithms and tasks, and improves the system's reconfiguration and evolvability.

Images