Autonomous unmanned platform

Abstract

The invention discloses an autonomous unmanned platform, and belongs to the field of artificial intelligence. The platform comprises a data acquisition layer, a data processing layer, an algorithm execution layer, a communication interaction layer and a system management layer. The data acquisition layer is used for data acquisition; the data processing layer is used for receiving and preprocessing data; the algorithm execution layer is used for inputting the preprocessed data into a corresponding algorithm to obtain an algorithm execution result, and monitoring and early warning the autonomous unmanned platform according to the algorithm execution result; the communication interaction layer is used for establishing communication connection among the data acquisition layer, the data processing layer, the algorithm execution layer and the system management layer; and the system management layer is used for monitoring the operation state of each layer of the platform and performing intelligent management on the platform according to the operation state. According to the invention, the platform architecture which can flexibly reconstruct, efficiently process data and execute a complex algorithm is realized, and the diversified requirements of different task scenes on the autonomous unmanned platform are met.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of artificial intelligence, and in particular, the present application relates to an autonomous unmanned platform. BACKGROUND

[0002] With the rapid development of artificial intelligence and automation technology, autonomous unmanned platforms have been widely used in military, industrial, civil and other fields. In practical applications, different task scenarios have different functional requirements for autonomous unmanned platforms. For example, environmental monitoring tasks require platforms to have efficient data collection and analysis capabilities, while target search tasks focus more on image recognition and intelligent decision-making capabilities.

[0003] Traditional autonomous unmanned platform software architecture is usually designed and developed for specific tasks, with relatively fixed functions, lacking flexibility and scalability. Once the task requirements change, the entire software system often needs to be redesigned and developed, which not only increases development costs and time, but also reduces software reusability.

[0004] In addition, existing software architectures have low computational resource utilization efficiency and slow algorithm execution speed when implementing complex algorithms such as FFT, neural networks, genetic algorithms, and deep learning algorithms, making it difficult to meet real-time monitoring and early warning requirements. At the same time, the interconnection between computing nodes in the software architecture is not flexible enough, and the topology structure cannot be quickly adjusted according to task requirements, limiting the performance improvement of autonomous unmanned platforms.

[0005] Therefore, the present application proposes an autonomous unmanned platform. SUMMARY

[0006] The present application aims to overcome the shortcomings of the prior art and proposes an autonomous unmanned platform to achieve the following objectives: a software architecture that can be flexibly reconfigured, efficiently process data, and execute complex algorithms, and adapt to the diverse needs of different task scenarios for autonomous unmanned platforms.

[0007] To achieve the above objectives, the technical solution adopted by the present application is as follows: an autonomous unmanned platform, the platform comprising a data collection layer, a data processing layer, an algorithm execution layer, a communication interaction layer, and a system management layer, wherein:

[0008] The data collection layer comprises a plurality of data collection modules for data collection and sending to the data processing layer;

[0009] The data processing layer is used to receive and preprocess the data before sending it to the algorithm execution layer;

[0010] The algorithm execution layer includes a plurality of algorithm plug-ins, which are configured to input the preprocessed data into corresponding algorithms to obtain algorithm execution results, and perform monitoring and early warning of the autonomous unmanned platform according to the algorithm execution results.

[0011] The communication interaction layer is configured to establish a communication connection among the data acquisition layer, the data processing layer, the algorithm execution layer, and the system management layer.

[0012] The system management layer is configured to monitor running states of the layers of the platform, and perform intelligent management of the platform according to the running states.

[0013] Preferably, each data acquisition module is connected to a corresponding sensor or external device through a standardized data acquisition interface, and is configured to send data acquired by the corresponding sensor or external device to the data processing layer.

[0014] Preferably, the sensor or external device includes a sonar, an inertial measurement unit (IMU), a millimeter wave radar, a depth sounder, a differential GPS, a laser radar, and a camera.

[0015] Preferably, the preprocessing operation includes data cleaning, data noise reduction, filtering, and feature extraction.

[0016] Preferably, each of the plurality of algorithm plug-ins is configured with a different algorithm, and is configured to analyze data input by the data processing layer through the built-in algorithm to obtain corresponding algorithm execution results, and perform monitoring and early warning of the autonomous unmanned platform according to the algorithm execution results.

[0017] Preferably, the algorithm execution layer uses an optimized algorithm execution technology when executing algorithms, and the optimized algorithm execution technology includes GPU acceleration and distributed computing.

[0018] Preferably, the communication interaction layer provides a plurality of communication protocols, including TCP / IP, UDP, and CAN bus.

[0019] Preferably, the system management layer monitors running states of the layers of the platform, including resource occupation, task execution progress, and system performance indicators.

[0020] Preferably, the system management layer performs intelligent management of the platform according to the running states, including:

[0021] determining whether a task executed by the current platform has changed;

[0022] If the task has changed, the system management layer performs platform architecture reconstruction according to the task change, including adding / removing data acquisition modules, adding / removing algorithm plug-ins, adjusting connection relationships among computing nodes in the data acquisition layer, the data processing layer, and the algorithm execution layer, and reallocating resources.

[0023] Preferably, the system management layer comprises a user interface for platform configuration and management according to user instructions.

[0024] The technical effects of the present application are:

[0025] (1) Flexibility and scalability: The platform can be flexibly configured and extended according to task requirements, and the topology structure can be dynamically adjusted by adopting a layered modular design and a reconfigurable mechanism. Users do not need to redevelop the entire software system, but only need to perform simple configuration and module addition or deletion operations to quickly adapt to different task scenarios, greatly improving the reusability of the software and reducing development costs and time.

[0026] (2) Efficient data processing and algorithm execution: The data processing layer adopts multi-threading and parallel computing technology, and the algorithm execution layer is optimized in combination with hardware characteristics, which can quickly process a large amount of data and efficiently execute complex algorithms such as FFT, neural network, genetic algorithm and deep learning algorithm, meeting the real-time monitoring and early warning requirements of the autonomous unmanned platform and improving the intelligent level and decision-making ability of the platform.

[0027] (3) Reliable communication and data transmission: The communication interaction layer realizes the interconnection of computing nodes through high-speed standardized buses and protocols, adopts adaptive communication strategies and data encryption technology to ensure fast and stable data transmission between different computing nodes, and guarantees the security and reliability of data, providing a solid foundation for the collaborative work between the layers of the software architecture.

[0028] (4) Intelligent system management: The system management layer realizes functions such as resource allocation, task scheduling, system monitoring, fault diagnosis and recovery, can reasonably utilize system resources, ensure timely execution of critical tasks, quickly diagnose and solve system faults, improve the stability and reliability of the software system, reduce manual intervention, and reduce operation and maintenance costs. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 An autonomous unmanned platform structure schematic diagram is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0030] The specific embodiments of the present application will be further described in detail below with reference to the drawings, and the purpose is to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present application, and to facilitate its implementation. It should be noted that the words "first", "second" and the like in the present application are only for the convenience of describing the technical solutions and used as a distinction between components, and the corresponding component configurations may be the same or different, which does not limit the present application. In order to make the technical solution of the present application more clear, the present application is explained and described by the following embodiments.

[0031] The embodiment provides an autonomous unmanned platform, which has a framework capable of realizing flexible reconstruction, efficiently processing data and executing complex algorithms, and can adapt to diversified requirements of different task scenarios for the autonomous unmanned platform.

[0032] It is stated in the foregoing that, when the autonomous unmanned platform is applied, hardware environment construction is required first. According to actual requirements of the autonomous unmanned platform, appropriate hardware devices are selected, including processors (such as high-performance CPUs, GPUs), storage devices (such as solid-state disks), sensors (such as various environmental sensors and image sensors), communication modules (such as wireless network cards and CAN bus controllers), and the like, which are reasonably installed on the unmanned platform, so as to ensure stable connection between the hardware devices and meet the requirements of the software framework on hardware resources. After the hardware environment construction is completed, the construction of the software framework can be performed.

[0033] As shown in Figure 1 The autonomous unmanned platform of the embodiment comprises a data acquisition layer, a data processing layer, an algorithm execution layer, a communication interaction layer and a system management layer, and each layer is provided with a standardized interface for connection between layers, so as to ensure smooth data interaction and normal function calling. Among them:

[0034] The data acquisition layer comprises a plurality of data acquisition modules, which are used for data acquisition and sending to the data processing layer;

[0035] The data processing layer is used for receiving and preprocessing the data and sending the preprocessed data to the algorithm execution layer;

[0036] The algorithm execution layer comprises a plurality of algorithm plug-ins, which are used for inputting the preprocessed data into corresponding algorithms to obtain algorithm execution results, and performing monitoring and early warning of the autonomous unmanned platform according to the algorithm execution results;

[0037] The communication interaction layer is used for establishing communication connection between the data acquisition layer, the data processing layer, the algorithm execution layer and the system management layer;

[0038] The system management layer is used for monitoring the running state of each layer of the platform, and intelligently managing the platform according to the running state.

[0039] Specifically, each data acquisition module of the embodiment is connected to a corresponding sensor or external device through a standardized data acquisition interface, and is used for sending the data acquired by the corresponding sensor or external device to the data processing layer.

[0040] The unmanned autonomous platform of the embodiment is applied to ship navigation, and the corresponding sensors or external devices used according to actual task requirements include sonar, inertial measurement unit (IMU), millimeter wave radar, depth sounder, differential GPS, laser radar and camera. In specific implementation, the sensors or external devices can be flexibly selected according to actual conditions. After preliminary format conversion and verification of the collected data, the data is transmitted to the data processing layer.

[0041] After the data processing layer receives the data, the data is subjected to data cleaning, data noise reduction, filtering, feature extraction and other preprocessing operations by using multi-threading and parallel computing technology, so as to provide the processed data for the algorithm execution layer.

[0042] In a preferred embodiment of the present application, the data processing layer is also modularized, that is, the data processing layer includes a plurality of preprocessing modules, each of which is used to execute at least one preprocessing operation. In specific implementation, according to actual task requirements, the required preprocessing module can be flexibly selected to join the data processing task of the data processing layer, so as to reduce unnecessary operations, optimize system resources and speed up work efficiency.

[0043] In the algorithm execution layer, the plurality of algorithm plug-ins of the embodiment are each configured with different algorithms, which are used to analyze the data input by the data processing layer by using the built-in algorithms to obtain corresponding algorithm execution results. In specific implementation, the required task algorithm plug-in can be selected according to actual task requirements. For example, the FFT algorithm plug-in is selected when signal analysis is performed, and the neural network or deep learning algorithm plug-in is selected when target recognition is performed. At the same time, according to the algorithm execution result, the autonomous unmanned platform is monitored and warned. For example, when the algorithm execution result shows that abnormal data or target is detected, the autonomous unmanned platform triggers the warning mechanism and sends alarm information to the user.

[0044] Further, in order to speed up algorithm execution and improve work efficiency, the algorithm execution layer of the embodiment uses optimized algorithm execution technology when executing algorithms, and the optimized algorithm execution technology includes GPU acceleration, distributed computing and the like.

[0045] According to the above, it can be found that the data acquisition layer, the data processing layer, and the algorithm execution layer of the embodiment include a plurality of computing nodes, including each data acquisition module and each algorithm plug-in, and in the preferred embodiment, each preprocessing module. It is the interconnection work between these computing nodes that realizes the task execution of the autonomous unmanned platform, and the interconnection work between the computing nodes relies on the data communication between the computing nodes. In order to improve the stability and efficiency of communication, the communication interaction layer of the embodiment establishes the connection between the computing nodes through a high-speed standardized bus and protocol, automatically selects a suitable communication protocol according to the network environment and task requirements, and configures the corresponding communication parameters. Common communication protocols include TCP / IP, UDP, CAN bus, etc. Further, in the data transmission process, the communication interaction layer adopts an adaptive communication strategy, that is, through indicators such as signal strength, packet loss rate, and delay, the network state is perceived, and according to a preset rule or a machine learning algorithm, the data transmission rate, retransmission mechanism, etc. are dynamically adjusted to ensure that data is quickly and stably transmitted between different computing nodes, for example, the AMC (Adaptive Modulation and Coding) technology commonly used in 4G / 5G networks. At the same time, according to the task requirements, the communication interaction layer can also encrypt the transmitted data to ensure the security of the data and prevent data leakage and tampering.

[0046] In addition, the communication interaction layer can also send the working data of the data acquisition layer, the data processing layer, and the algorithm execution layer (for example, the start-stop situation of each computing node, the resources occupied, etc.), as well as its own working data (for example, the communication protocol used between each computing node, the rate of communication, the delay, etc.) to the system management layer, thereby meeting the monitoring needs of the system management layer. In the embodiment, the running states of the platform layers monitored by the system management layer include resource occupation, task execution progress, system performance indicators, etc., so as to intelligently manage the autonomous unmanned platform according to these running states.

[0047] For example, the system management layer reasonably allocates computing resources, storage resources, and communication resources according to the needs of each module through a dynamic resource allocation algorithm (such as a greedy algorithm, a maximum-minimum fairness algorithm), such as when the algorithm execution layer needs more computing resources, the system management layer allocates idle CPU or GPU resources to this layer. Using a task scheduling algorithm (such as a priority scheduling algorithm), different tasks are prioritized and scheduled to ensure that critical tasks are executed first. When the system detects a fault, the system management layer quickly diagnoses the fault, locates the cause of the fault, and takes appropriate recovery measures, such as restarting the faulty module, switching to a backup module, etc., to ensure the stable operation of the software system.

[0048] In addition, the system management layer of the embodiment includes a user interface for platform configuration and management according to user instructions, thereby facilitating user direct configuration and management of the autonomous unmanned platform, such as adjusting data acquisition parameters, replacing algorithm plug-ins, monitoring system operation logs, and the like.

[0049] Meanwhile, the system management layer of the embodiment performs intelligent management of the platform according to the running state, and further includes:

[0050] determining whether the task performed by the current platform has changed; if the task has changed, the system management layer performs platform architecture reconstruction according to the task change, including adding / removing data acquisition modules, adding / removing algorithm plug-ins, adjusting the connection relationship between the computing nodes in the data acquisition layer, the data processing layer, and the algorithm execution layer, and re-allocating resources. For example, during task execution, when the task demand changes, new monitoring functions need to be added, at which time the user can add corresponding data acquisition modules, preprocessing modules, and algorithm plug-ins in the system management layer, adjust the connection relationship between the computing nodes, re-allocate resources, and realize rapid reconstruction of the software architecture.

[0051] In summary, the present application establishes a reconfigurable autonomous unmanned platform by constructing a layered and modular architecture, including five layers of data acquisition, data processing, algorithm execution, communication interaction, and system management, and realizes low coupling through standardized interfaces, supports dynamic addition and deletion of functional modules and flexible recombination of topological structures. The autonomous unmanned platform has the advantages of multi-type sensor plug-and-play, multi-thread parallel processing, plug-in algorithm integration, and adaptive communication strategy, solves the problems of fixed function, poor scalability, and low reliability of traditional architectures, and realizes flexible adaptation and stable operation of the unmanned platform in different task scenarios.

[0052] The present application has been described above with reference to the drawings. Obviously, the specific implementation of the present application is not limited to the above-described manner. Any non-essential improvement or direct application of the above-described concept and technical solution of the present application to other occasions is within the scope of protection of the present application.

Claims

1. An autonomous unmanned platform, characterized by: The platform includes a data acquisition layer, a data processing layer, an algorithm execution layer, a communication interaction layer, and a system management layer, wherein: The data acquisition layer includes multiple data acquisition modules for collecting data and sending it to the data processing layer; The data processing layer is used to receive and pre-process the data and then send it to the algorithm execution layer; The algorithm execution layer includes multiple algorithm plug-ins, which are used to input the pre-processed data into the corresponding algorithm to obtain the algorithm execution results, and perform monitoring and early warning of the autonomous unmanned platform based on the algorithm execution results; The communication interaction layer is used to establish communication connections between the data acquisition layer, data processing layer, algorithm execution layer, and system management layer; The system management layer is used to monitor the operating status of each layer of the platform and perform intelligent management of the platform according to the operating status.

2. The autonomous unmanned platform according to claim 1, characterized in that: Each data acquisition module is connected to a corresponding sensor or external device via a standardized data acquisition interface, and is used to collect data through the corresponding sensor or external device and then send the data to the data processing layer.

3. The autonomous unmanned platform according to claim 2, characterized in that: The sensors or external devices include sonar, inertial measurement unit (IMU), millimeter wave radar, depth sounder, differential GPS, lidar, and camera.

4. The autonomous unmanned platform according to claim 1, characterized in that: The preprocessing operations include data cleaning, data noise reduction, filtering, and feature extraction.

5. The autonomous unmanned platform according to claim 1, characterized in that: The multiple algorithm plug-ins are all configured with different algorithms, which are used to analyze the data input by the data processing layer through the built-in algorithm to obtain the corresponding algorithm execution results, and perform monitoring and early warning of the autonomous unmanned platform based on the algorithm execution results.

6. An autonomous unmanned platform according to claim 1 or 5, characterized in that: The algorithm execution layer adopts optimized algorithm execution technology when executing the algorithm, and the optimized algorithm execution technology includes GPU acceleration and distributed computing.

7. The autonomous unmanned platform according to claim 1, characterized in that: The communication interaction layer provides multiple communication protocols, including TCP / IP, UDP, and CAN bus.

8. The autonomous unmanned platform according to claim 1, characterized in that: The operating status of each layer of the platform monitored by the system management layer includes resource usage, task execution progress, and system performance indicators.

9. The autonomous unmanned platform according to claim 1, characterized in that: The system management layer performs intelligent platform management according to the operating status, including: Determine whether the task executed by the current platform has changed; If any changes occur, the system management layer will reconstruct the platform architecture according to the task changes, including: adding / deleting data acquisition modules; adding / deleting algorithm plug-ins; adjusting the connection relationship between each computing node in the data acquisition layer, data processing layer, and algorithm execution layer; and reallocating resources.

10. The autonomous unmanned platform according to claim 1, characterized in that: The system management layer includes a user interface for configuring and managing the platform according to user instructions.

Images