CN112558624B - Spacecraft autonomous mission planning verification deployment integrated intelligent computing system - Google Patents

Abstract

The invention discloses an integrated intelligent computing system for autonomous mission planning, verification and deployment of a spacecraft, which comprises the following components: a high-performance mobile workstation unit and a high-performance heterogeneous computing unit; the high-performance mobile workstation unit downloads the compiled target code to the high-performance heterogeneous computing unit, and the compiled target code runs in the high-performance heterogeneous computing unit; a bus interface on the high-performance heterogeneous computing unit acquires the state of the spacecraft and transmits the state of the spacecraft to the high-performance mobile workstation unit; the high-performance mobile workstation unit stores and interprets the state of the spacecraft to obtain a control instruction, and transmits the control instruction to the high-performance heterogeneous computing unit; and the high-performance heterogeneous computing unit outputs the control instruction to the spacecraft related subsystem. The invention can process large data volume, can meet the algorithm time and space complexity required by intelligent calculation algorithm, has parallel calculation capability and ensures real-time performance.

Description

An integrated intelligent computing system for spacecraft autonomous mission planning, verification and deployment

Technical field

The invention belongs to the technical field of intelligent computing systems, and in particular relates to an integrated intelligent computing system for spacecraft autonomous mission planning, verification and deployment.

Background technique

The implementation of autonomous mission planning technology is different from the traditional spacecraft development model. On the one hand, autonomous mission planning requires the acquisition and accumulation of large-scale, multi-mode own on-orbit status data, and the completion of near-real-time, high-performance intelligent computing on large-scale data, thereby Obtain planning results; on the other hand, the above-mentioned on-orbit state model is huge and has many trigger modes, making it difficult to rely on traditional simulation methods for effective simulation and verification traversal, and cannot meet the timeliness of the mission cycle. Therefore, the spaceborne system is required to be able to adapt to verification, The deployment and testing process are seamless and iterative. However, the current spaceborne computing system system is difficult to meet the requirements from the above two aspects.

Contents of the invention

The technical problem solved by the present invention is to overcome the shortcomings of the existing technology and provide an integrated intelligent computing system for spacecraft autonomous mission planning, verification and deployment, which can process large amounts of data and satisfy the algorithm time and time required by intelligent computing algorithms. Space complexity, parallel computing capabilities, and real-time performance. At the same time, in terms of usage mode, it must be able to have both verification and deployment functions at the same time, and be able to seamlessly connect in the process of technical verification, deployment testing, correction and verification.

The object of the present invention is achieved through the following technical solution: an integrated intelligent computing system for spacecraft autonomous mission planning verification and deployment, including: a high-performance mobile workstation unit and a high-performance heterogeneous computing unit; wherein the high-performance mobile workstation unit will The compiled target code is downloaded to the high-performance heterogeneous computing unit, and the compiled target code runs on the high-performance heterogeneous computing unit; the bus interface on the high-performance heterogeneous computing unit obtains the spacecraft status and transmits the spacecraft status to The high-performance mobile workstation unit; the high-performance mobile workstation unit stores and interprets the status of the spacecraft to obtain control instructions, and transmits the control instructions to the high-performance heterogeneous computing unit; the high-performance heterogeneous computing unit transmits the control instructions to the spacecraft-related Sub-system output.

In the above-mentioned spacecraft autonomous mission planning verification deployment integrated intelligent computing system, the bus interface on the high-performance heterogeneous computing unit obtains the spacecraft status, and the high-performance heterogeneous computing unit calculates the control instructions based on the spacecraft status. The high-performance heterogeneous computing unit The bus system of the computing unit outputs control instructions to the relevant subsystems of the spacecraft; the high-performance mobile workstation unit communicates with the high-performance heterogeneous computing unit or spacecraft subsystem through its wifi, wired Ethernet or thunderblot bus for monitoring Data when high-performance heterogeneous computing units independently run flight control programs.

In the above-mentioned spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system, the high-performance heterogeneous computing unit includes an image processor, a first central processing unit, a second central processing unit, a programmable gate array, a memory, and a high-speed internal bus , 1553B bus, 100/Gigabit Ethernet and data conversion and acquisition interface; among them, the first central processor obtains the spacecraft platform status data through the high-speed internal bus, 1553B bus, 100/Gigabit Ethernet and data conversion and acquisition interface; wherein , the spacecraft platform status data includes graphic image data, large-scale time series data and event data; the first central processor transmits the graphic image data to the image processor through a high-speed internal bus, and the image processor performs operations on the graphic image data, and then The image processor returns the operation results to the first central processor through the high-speed internal bus; the first central processor transmits large-scale timing data to the programmable gate array for processing through the high-speed internal bus, and then the programmable gate array processes The results are returned to the first central processor through the high-speed internal bus; the first central processor transmits the event data to the second central processor through the high-speed internal bus for processing, and then the second central processor transmits the processing results through the high-speed internal bus Return to the first central processor; the first central processor completes information processing based on the processing results returned by the second central processor, the operation results returned by the image processor, and the processing results returned by the programmable gate array.

Compared with the prior art, the present invention has the following beneficial effects:

(1) Through the combined design of high-performance computing units and heterogeneous computing units, the present invention combines various functions such as human-computer interaction verification links and quick access system unique interfaces, and designs different working modes in separate and combined modes. , achieves the effect of adapting to different usage requirements in the verification stage and deployment operation stage, and lays the foundation for being suitable for independent intervention systems to serve as autonomous operation planning machines.

(2) Through high-performance general-purpose computing units, the present invention achieves the effects of supporting human-computer interaction, synchronizing ground research and development environments, deploying development systems, completing large-scale computing tasks such as training intelligent control or planning models, and completing large-scale test data storage.

(3) Through the scalable I/O module of the high-performance heterogeneous computing unit, the present invention can expand the effects of commonly used spacecraft bus forms, including 1553B bus, GigE bus, WIFI, etc., and can be integrated with other subsystems of the spacecraft. Perform data communication and interact with high-performance general-purpose computing units.

(4) Through the combined design of high-performance computing units and heterogeneous computing units, the present invention achieves the effect of supporting different application modes, and can complete independent planning and intelligent control tasks jointly or separately. In the joint mode, the heterogeneous computing unit can be used as an accelerator of a high-performance general-purpose computing unit and as an interface module such as the external system 1553B bus; in a discrete mode, the heterogeneous computing unit can be used as a high-performance spaceborne computer on which to deploy After training and verification, the control application is planned and actually interacts with sensors and other subsystems to deploy, run, and verify algorithms and programs.

Description of the drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are for the purpose of illustrating preferred embodiments only and are not to be construed as limiting the invention. Also throughout the drawings, the same reference characters are used to designate the same components. In the attached picture:

Figure 1 is a schematic diagram of the information flow of the operating status of the spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system provided by an embodiment of the present invention;

Figure 2 is a schematic diagram of the use of three operating modes of the spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system provided by the embodiment of the present invention;

Figure 3 is a schematic diagram of the software composition of a high-performance mobile workstation system provided by an embodiment of the present invention;

Figure 4 is a schematic diagram of the composition of a high-performance heterogeneous computing unit provided by an embodiment of the present invention;

Figure 5 is a schematic diagram of the software composition of a high-performance heterogeneous computing unit system provided by an embodiment of the present invention.

Detailed ways

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a thorough understanding of the disclosure, and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, as long as there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

Figure 1 is a schematic diagram of the information flow of the operating status of the spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system provided by an embodiment of the present invention. As shown in Figure 1, the spacecraft autonomous mission planning verification deployment integrated intelligent computing system includes: a high-performance mobile workstation unit and a high-performance heterogeneous computing unit; wherein the high-performance mobile workstation unit downloads the compiled target code Installed into the high-performance heterogeneous computing unit, the compiled target code is run on the high-performance heterogeneous computing unit; the bus interface on the high-performance heterogeneous computing unit obtains the spacecraft status and transmits the spacecraft status to the high-performance mobile workstation unit; The high-performance mobile workstation unit stores and interprets the status of the spacecraft to obtain control instructions, and transmits the control instructions to the high-performance heterogeneous computing unit; the high-performance heterogeneous computing unit outputs the control instructions to the relevant subsystems of the spacecraft.

The bus interface on the high-performance heterogeneous computing unit obtains the spacecraft status, the high-performance heterogeneous computing unit calculates the control instructions based on the spacecraft status, and the bus system of the high-performance heterogeneous computing unit outputs control instructions to the relevant subsystems of the spacecraft; The high-performance mobile workstation unit communicates with the high-performance heterogeneous computing unit or spacecraft subsystem through its WiFi, wired Ethernet or thunderblot bus, and is used to monitor data when the high-performance heterogeneous computing unit independently runs the flight control program.

It is composed of a high-performance mobile workstation unit and a high-performance heterogeneous computing unit, which are connected through a mechanical interface and a high-speed bus interface. In different application modes, independent planning and intelligent control tasks can be completed jointly or separately to achieve R&D test mode, R&D verification mode, and deployment operation mode.

Among them, when the system is in R&D test mode, it mainly uses development environments installed in high-performance mobile workstation units, such as MATLAB, Visual Studio, LabVIEW, etc., to conduct research and development or simulation of intelligent algorithms, independent program prototypes, and independent control models; at the same time, high-performance The heterogeneous computing unit plays the role of an extended interface between the co-processor and external systems. It is responsible for assisting the mobile workstation unit in accelerating calculations and providing data communication capabilities. The workstation unit provides a human-machine interface for astronauts to intervene in operations.

Among them, when the system is in R&D verification mode, the compiled target code can be downloaded to the heterogeneous computing unit using the development and deployment program in the high-performance mobile workstation unit. The target code runs in the heterogeneous computing unit while the mobile workstation unit The executable monitoring software communicates with the heterogeneous computing unit, obtains the spacecraft status and test data through the bus interface on the high-performance heterogeneous computing unit, and the high-performance mobile workstation unit stores, transmits, and interprets the test results, and obtains control instructions. , the high-performance mobile workstation unit outputs the output control instructions to the relevant subsystems of the spacecraft through the bus system of the high-performance heterogeneous computing unit. The workstation unit provides a human-machine interface for astronauts to intervene in operations.

Among them, when the system is in deployment operation mode, the high-performance heterogeneous computing unit and the high-performance mobile workstation unit are in a separate state. The compiled and optimized actual autonomous flight program has been deployed in the high-performance heterogeneous computing unit. The heterogeneous computing unit communicates with each subsystem of the spacecraft through the data bus to perform actual autonomous planning and intelligent control. Control instructions are output to the spacecraft-related subsystems through the bus system of high-performance heterogeneous computing units. In this case, the high-performance mobile workstation unit can communicate with the heterogeneous computing units or spacecraft through the wifi or wired Ethernet or thunderblot bus on it. The computer communicates with subsystems and is used to monitor data when heterogeneous computing units independently run flight control programs.

The high-performance mobile workstation unit is a high-performance computing platform and the main human-computer interaction interface of the system. It can be supported by the IA64 Intel architecture platform. A virtual machine system can be deployed on it. Windows or Linux operating systems can be flexibly deployed and can run in conjunction with the ground R&D environment. Consistent, development environments such as Matlab, Visual Studio, GCC, etc. can be synchronized with the ground test environment; support development libraries such as Caffe, Tensorflow, CUDA, etc., for large-scale data fusion calculations, machine learning, and intelligent control model training and other test verification work; it has a large-capacity solid-state drive that can store process data. The high-performance mobile workstation communicates data with the high-performance heterogeneous computing unit through the Thunderbolt bus, with a bandwidth of 40Gbps, thereby expanding computing acceleration resources and expanding I/O resources (such as 1553B, AD, DA, etc.), and at the same time, the calculated and verified The autonomous flight control program is deployed to the heterogeneous computing unit through the development environment, and the heterogeneous computing unit independently undertakes the task. The workstation serves as a monitoring host computer and communicates with the heterogeneous computing unit or data management subsystem through wifi or wired Ethernet or thunderblot bus. The telemetry subsystem communicates and monitors data when the heterogeneous computing units independently run the flight control program.

The high-performance heterogeneous computing unit includes an image processor, a first central processing unit 1, a second central processing unit 2, a programmable gate array, a memory, a high-speed internal bus, a 1553B bus, a 100/gigabit Ethernet, and a data conversion and acquisition interface. ;in,

The high-performance mobile workstation unit runs a compilation and synthesis software combination. After the compilation and synthesis software combination compiles and synthesizes the program, it will generate CPU machine code files that run on the CPU 1 and 2 respectively. The sequential logic files of the programming gate array and the machine code files running on the image processor are then transmitted to the high-performance heterogeneous computer through the thunderbolt bus. The unit stores the above files in the memory respectively. The storage location of the memory corresponds to the central processing unit. The storage area of the CPU 1, the storage area of the CPU 2, the programmable gate array storage area and the image processor storage area. After each component in the high-performance heterogeneous computing unit is powered on or restarted, files are loaded from the corresponding areas.

The central processor 1 runs the program in its storage area and completes the 1553 bus interface chip, 100/gigabit Ethernet, and data conversion and acquisition interface circuits mounted on the central processor 1 to obtain the spacecraft platform status data.

After acquiring the data, the central processing unit 1 transmits the graphic image data to the image processor through the high-speed internal bus. The image processor performs operations on the image, and then the image processor returns the operation results to the central processor 1 through the high-speed internal bus; The large-scale timing data is transmitted to the programmable gate array through the high-speed internal bus for processing, and then the programmable gate array sends the operation results to the central processor 1 through the high-speed internal bus; the event data is transmitted to the central processing unit through the high-speed internal bus. 2 performs processing, and then the central processor 2 returns the operation results to the central processor 1 through the high-speed internal bus; finally, the central processor 1 completes the synthesis of the data returned by the central processor 2, the image processor and the programmable gate array calculation to complete information processing.

The high-performance heterogeneous computing unit has central processing unit 1/2, high-performance image processor, large-scale FPGA resources and scalable I/O interface capabilities. In the joint working mode, it is combined with the workstation unit through the mechanical interface, electrically connected through the Thunderbolt bus, and performs data communication with the mobile workstation. This mode is used as a parallel computing resource. In independent working mode, it can be used as a future high-performance heterogeneous spaceborne computer, and can deploy autonomous systems with high concurrency, large-scale data, and complex parallel real-time characteristics. Among them, the central processing unit is used to deploy the runtime library, file system and driver used to run the program; the high-performance image processor module is mainly used for calculation acceleration and large-scale high-performance computing. For example: the experimental verification process of the algorithm can be performed in high-performance Use the DIGITS interface to train the deep learning model on the mobile workstation, and then use the high-performance inference engine TensorRT equipped with JetPack to deploy the trained model to a high-performance image processor. The high-performance image processor is used to operate in cooperation with the central processor 1 to complete calculation acceleration and large-scale high-performance operations; the field programmable gate array is used for specific operation logic, specific acceleration operation units, etc., and is managed and loaded by the central processor 2 in a unified manner Timing is a large-scale, high-performance field programmable gate array product. Among them, the scalable I/O module can expand the I/O interface resources required for tasks, especially the bus I/O required for communication with external systems, such as the 1553B interface and the high-performance Gigabit Ethernet interface with customizable protocols. , WIFI can realize data communication between mobile workstations and heterogeneous computing units through point-to-point, high-speed, high-bandwidth and extended I/O.

As shown in Figure 1, the spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system consists of a high-performance general computing unit and a high-performance heterogeneous computing unit, which are connected through a mechanical interface and a high-speed bus interface. Run autonomous intelligent control support software on the system, as shown in Figure 1. Driven by software, the spacecraft autonomous mission planning, verification and deployment integrated intelligent computing system can expand the bus that meets the requirements of the spacecraft through high-performance heterogeneous computing units, complete data communication with each subsystem of the spacecraft, and obtain the spacecraft The system status of sensors, spacecraft actuators, and related subsystems is sent to the autonomous planning and intelligent control system for calculation and autonomous mission planning, and then the planning results are converted into spacecraft-related command sequences and processed through the bus system. The device-related subsystem outputs control signals. In addition, the integrated intelligent computing system for spacecraft autonomous mission planning, verification and deployment can receive astronaut instructions through high-performance general-purpose computing units, achieve human-computer interaction, synchronize ground research and development environments, deploy development systems, and complete large-scale training of intelligent control or planning models, etc. Compute tasks and complete large-scale test data storage.

As shown in Figure 2, high-performance general computing units and high-performance heterogeneous computing units can be connected or separated to complete independent planning and intelligent control tasks, and can cope with three application modes:

① R&D test mode: The system is in R&D test mode, mainly using the development environment installed in the high-performance general-purpose computing unit to conduct research and development or simulation of intelligent algorithms, independent program prototypes, and independent control models; at the same time, the high-performance heterogeneous computing unit is in co-processing It plays the role of an extended interface between the computer and external systems, and is responsible for assisting the high-performance general-purpose computing unit unit in accelerating calculations and providing data communication capabilities. The workstation unit provides a human-machine interface for astronauts to intervene in operations.

② R&D verification mode: When the system is in R&D verification mode, the high-performance heterogeneous computing unit is regarded as a high-performance spaceborne computer. Using the development and deployment program in the high-performance general-purpose computing unit, the compiled target code can be downloaded to the heterogeneous computing unit. The target code runs in the heterogeneous computing unit. At the same time, the high-performance general-purpose computing unit can run monitoring software and heterogeneous computing units. The computing unit communicates with the computer to obtain test data, and stores, transmits, and interprets test results. The workstation unit provides a human-machine interface for astronauts to intervene.

③ Deployment operation mode: When the system is in the deployment operation mode, the high-performance heterogeneous computing unit and the high-performance space-borne computer are in a separate state. The compiled and optimized actual autonomous flight program has been deployed in the heterogeneous computing unit. The heterogeneous computing unit communicates with each subsystem through the data bus to perform actual autonomous planning and intelligent control.

As shown in Figure 3, high-performance mobile workstations are mainly used to deploy R&D environments online and support test verification tasks. They are characterized by running non-real-time, large-scale high-performance computing tasks, deploying development environments required for test and verification processes, and running heterogeneous The program deployment environment required by the computing platform.

Supported by the IA64 Intel architecture platform, a virtual machine system is deployed on it to flexibly deploy a shelf-style mature operating system; a development library consistent with the ground R&D platform is deployed to synchronize with the ground test environment for large-scale data fusion computing, machine Learning, intelligent control model training and other experimental verification work; equipped with large-capacity memory, which can store process data; perform data communication with high-performance heterogeneous computing units through internal high-speed buses, thereby expanding computing acceleration resources and expanding I through heterogeneous computing units /O resources (such as 1553B, AD, DA, etc.); deploy the autonomous flight control program after calculation and verification to the heterogeneous computing unit through the development environment, and the heterogeneous computing unit independently undertakes the task; communicate with the heterogeneous computing unit through a high-speed bus And each sub-system communicates and monitors the data when the heterogeneous computing units independently run the flight control program; the high-performance mobile workstation is mainly divided into two working states, namely the on-orbit operating state and the ground collaboration state.

On-orbit operating status

In the orbital operating state, high-performance mobile workstations mainly run applications and need to be equipped with basic items including operating systems, files and storage;

Applications: Deploy test programs, verify programs, run non-real-time autonomous programs.

Operating system: It mainly deploys the operating system that can be used to support the necessary runtime libraries and operating environment for running application software. This design uses Windows system as the operating system software.

Files and Storage: Deploy a database for data storage.

Ground coordination status

In the ground collaboration state, in addition to the above-mentioned basic items required in the on-orbit operation stage, development and compilation software also needs to be installed, including virtualization, development environment, runtime library, program development framework, etc.

Virtualization: The system deploys virtualized basic software to run a wider range of ecological software systems required for research and development.

Development environment: Deploy the software platform required for R&D processes such as intelligent independent algorithm development and heterogeneous computing unit embedded program development.

Runtime library: deploys virtualization software, development environment software, storage systems, program frameworks, and support library software required for application runtime.

Program development framework: The development framework required to deploy artificial intelligence, machine vision, machine learning, etc.

As shown in Figure 4, the high-performance heterogeneous computing unit has high-performance GPU and large-scale FPGA resources and scalable I/O interface capabilities, and is composed of GPU, FPGA, and scalable I/O modules.

The scalable I/O module can expand the 1553B bus, GigE bus, WIFI and other interfaces for control and data exchange with external systems.

In the joint working mode, it is combined with the workstation unit through the mechanical interface, electrically connected through the high-speed internal bus, and performs data communication with the high-performance general computing unit. This mode is used as a parallel computing resource.

In independent working mode, it can be used as a future high-performance heterogeneous spaceborne computer, and can deploy autonomous systems with high concurrency, large-scale data, and complex parallel real-time characteristics.

As shown in Figure 5, the high-performance heterogeneous computing unit mainly performs high-performance real-time calculations and is used to run flight verification programs and real intelligent autonomous flight programs. The system software in the heterogeneous computing unit is divided into three parts, and the distribution of the three parts is completed by the development program running in the workstation unit.

Real-time applications are divided into CPU parts, GPU parts, and FPGA parts according to computing characteristics.

CPU part: The CPU is a multi-core ARM architecture A53 series, which runs a real-time Linux operating system, and the runtime library, file system and driver used to run the program are deployed on it.

GPU part: The program in the GPU part is a workstation development program. During the development or compilation process, the program parts that can be parallelized by a general-purpose GPU are identified and will be compiled into a specific format and run in collaboration with ARM.

FPGA part: The program of the FPGA part is designed during the development and design process of the independent intelligent flight program as specific operation logic, specific acceleration operation unit, IP soft core and other parts. After the program is compiled and synthesized through compilation and synthesis software combination, it is downloaded to In the FPGA configuration chip, ARM manages the loading sequence uniformly.

The three are deployed separately, and subsequent operations communicate through the PCIe bus for data synchronization.

Through the combined design of high-performance computing units and heterogeneous computing units, the present invention combines various functions such as human-computer interaction verification links and quick access system unique interfaces, and designs different working modes in separate and combined modes, achieving the goal of The effect of adapting to different usage requirements in the verification stage and deployment operation stage lays the foundation for being suitable for independent intervention systems to serve as autonomous operation planning machines.

By designing an interface mode that is software reconfigurable and capable of expansion, the present invention achieves the ability of the computing system to adapt to different spacecraft systems, thereby having the ability to respond to multiple categories of spacecraft systems, and carry out continuous autonomous mission planning related technology development and development on it. Extended utility capabilities.

Through high-performance general-purpose computing units, the present invention achieves the effects of supporting human-computer interaction, synchronizing ground research and development environments, deploying development systems, completing large-scale computing tasks such as training intelligent control or planning models, and completing large-scale test data storage.

Through the scalable I/O module of the high-performance heterogeneous computing unit, the present invention can expand the effects of commonly used spacecraft bus forms, including 1553B bus, GigE bus, WIFI, etc., and can perform data communication with other subsystems of the spacecraft. , while enabling data interaction with high-performance general-purpose computing units.

Through the combined design of high-performance computing units and heterogeneous computing units, the present invention achieves the effect of supporting different application modes, and can complete independent planning and intelligent control tasks jointly or separately. In the joint mode, the heterogeneous computing unit can be used as an accelerator of a high-performance general-purpose computing unit and as an interface module such as the external system 1553B bus; in a discrete mode, the heterogeneous computing unit can be used as a high-performance spaceborne computer on which to deploy After training and verification, the control application is planned and actually interacts with sensors and other subsystems to deploy, run, and verify algorithms and programs.

Although the present invention has been disclosed above in terms of preferred embodiments, they are not intended to limit the present invention. Any person skilled in the art can utilize the methods and technical contents disclosed above to improve the present invention without departing from the spirit and scope of the present invention. Possible changes and modifications are made to the technical solution. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention, all belong to the technical solution of the present invention. protected range.

Claims (2)

1. An integrated intelligent computing system for autonomous mission planning, verification and deployment of a spacecraft is characterized by comprising: a high-performance mobile workstation unit and a high-performance heterogeneous computing unit; wherein,,

the high-performance mobile workstation unit downloads the compiled object code to the high-performance heterogeneous computing unit, and the compiled object code runs in the high-performance heterogeneous computing unit;

a bus interface on the high-performance heterogeneous computing unit acquires the state of the spacecraft and transmits the state of the spacecraft to the high-performance mobile workstation unit; the high-performance mobile workstation unit stores and interprets the state of the spacecraft to obtain a control instruction, and transmits the control instruction to the high-performance heterogeneous computing unit; the high-performance heterogeneous computing unit outputs a control instruction to the spacecraft related subsystem; wherein,,

the high-performance heterogeneous computing unit comprises an image processor, a first central processing unit, a second central processing unit, a programmable gate array, a memory, a high-speed internal bus, a 1553B bus, a hundred/gigabit Ethernet and a data conversion acquisition interface; wherein,,

the first central processing unit acquires the state data of the spacecraft platform through a high-speed internal bus, a 1553B bus, a hundred/gigabit Ethernet and a data conversion acquisition interface; the spacecraft platform state data comprise graphic image data, large-scale time sequence data and event data;

the first central processing unit transmits the graphic image data to the image processor through the high-speed internal bus, the image processor operates the graphic image data, and then the image processor returns the operation result to the first central processing unit through the high-speed internal bus;

the first central processing unit transmits the large-scale time sequence data to the programmable gate array for processing through the high-speed internal bus, and then the programmable gate array returns the processing result to the first central processing unit through the high-speed internal bus;

the first central processing unit transmits event data to the second central processing unit through the high-speed internal bus for processing, and then the second central processing unit returns a processing result to the first central processing unit through the high-speed internal bus; the first central processing unit completes information processing according to the processing result returned by the second central processing unit, the operation result returned by the image processor and the processing result returned by the programmable gate array;

the high-performance mobile workstation unit and the high-performance heterogeneous computing unit can be simultaneously and separately used for completing autonomous planning and intelligent control tasks, and the three application modes are treated:

(1) research and development test mode: the system is in a research and development test mode, and mainly utilizes a development environment installed in a high-performance mobile workstation unit to research and develop an intelligent algorithm and an autonomous program prototype or simulate the autonomous program prototype and an autonomous control model; meanwhile, the high-performance heterogeneous computing unit is in the role of an expansion interface between the coprocessor and an external system, and is responsible for assisting the high-performance mobile workstation unit in accelerating computation and providing data communication capacity, and the workstation unit provides a man-machine interface for intervention operation of astronauts;

(2) research and development verification mode: when the system is in a research and development verification mode, the high-performance heterogeneous computing unit is regarded as a high-performance spaceborne computer; the method comprises the steps that a development deployment program in a high-performance mobile workstation unit is utilized, compiled target codes can be downloaded to a high-performance heterogeneous computing unit, the target codes run in the high-performance heterogeneous computing unit, meanwhile, monitoring software run by the high-performance mobile workstation unit is communicated with the high-performance heterogeneous computing unit, test data are obtained, the test data are stored, transmitted and interpreted, and a human-computer interface is provided for a spaceman to intervene in operation;

(3) deployment of the operational mode: when the system is in a deployment operation mode, the high-performance heterogeneous computing unit and the high-performance mobile workstation unit are in a discrete state; the high-performance heterogeneous computing unit is provided with an actual autonomous flight program after compiling optimization, and performs data communication with each subsystem of the spacecraft through a data bus to perform actual autonomous planning and intelligent control.

2. The spacecraft autonomous mission planning verification deployment integrated intelligent computing system of claim 1, wherein: the bus interface on the high-performance heterogeneous computing unit acquires the state of the spacecraft, the high-performance heterogeneous computing unit calculates according to the state of the spacecraft to obtain a control instruction, and the bus system of the high-performance heterogeneous computing unit outputs the control instruction to the related sub-system of the spacecraft;

the high-performance mobile workstation unit communicates with the high-performance heterogeneous computing unit or the spacecraft subsystem through wifi, wired Ethernet or thunderbolt buses on the high-performance mobile workstation unit and is used for monitoring data when the high-performance heterogeneous computing unit independently runs the flight control program.