Search

CN-122018493-A - Unmanned ship control method

CN122018493ACN 122018493 ACN122018493 ACN 122018493ACN-122018493-A

Abstract

The invention relates to the technical field of unmanned ship control, in particular to a control method of an unmanned ship, which upgrades a sensing device of the unmanned ship so as to acquire more accurate environmental data. Aiming at the characteristics of unmanned ship control, a deep learning algorithm is designed and optimized, so that the deep learning algorithm can more effectively extract features from perception data and generate control instructions. Based on a deep learning algorithm, an automatic intervention mechanism is designed, and when the control of the unmanned ship is interfered by the external environment, the control strategy can be automatically adjusted to maintain the accuracy and stability of the control. And a real-time monitoring system is established to monitor and evaluate the control process of the unmanned ship in real time. By collecting and analyzing feedback information of the unmanned ship, the performance of the automatic intervention mechanism is continuously adjusted and optimized. The unmanned aerial vehicle effectively solves the problem caused by the lack of an effective automatic intervention mechanism, and improves the accuracy and stability of unmanned aerial vehicle control.

Inventors

  • Lai Wanfang

Assignees

  • 马达西奇飞机发动机厂(湖北)有限公司

Dates

Publication Date
20260512
Application Date
20240312

Claims (6)

  1. 1. The unmanned ship control method is characterized by comprising the following steps of: upgrading a sensing device of the unmanned ship to obtain environment sensing data; Designing and optimizing a deep learning algorithm to extract features from the perception data and generate a control instruction; based on a deep learning algorithm, designing an automatic intervention mechanism; A real-time monitoring system is established to monitor and evaluate the control process of the unmanned ship in real time; And testing and verifying the unmanned ship, and continuously optimizing and adjusting a deep learning algorithm and an automatic intervention mechanism according to the test result.
  2. 2. The method of claim 1, wherein the step of upgrading the unmanned ship's sensing device to obtain environmental sensing data further comprises: and introducing a multi-sensor fusion technology, and carrying out fusion processing on data collected by different sensors.
  3. 3. The unmanned ship control method of claim 2, wherein the deep learning algorithm is designed and optimized to extract features from the perceived data and to generate control instructions, the steps further comprising: collecting and preparing perception data; Selecting a deep learning model architecture according to task requirements and data characteristics; designing a network layer and an activation function, performing feature pre-training by using an unsupervised learning method, introducing an attention mechanism or a feature fusion technology, and enhancing the attention capability of an algorithm to key features; Designing an output layer; Model training is carried out by using the loss function and an optimization algorithm; and evaluating the trained model by using the verification set, analyzing the performance of the model, and optimizing the model according to the evaluation result.
  4. 4. A method of unmanned aerial vehicle steering as claimed in claim 3, wherein the automated intervention mechanism is designed based on a deep learning algorithm, the steps further comprising: collecting control data of the unmanned ship in various environments and conditions, marking the data, and determining which conditions need automatic intervention and the targets and expected results of the intervention; training a deep learning model by using the marked data, so that the deep learning model can identify a scene needing automatic intervention and forecast a corresponding control instruction; According to the result of model prediction, designing an automatic intervention strategy; The automatic intervention mechanism is integrated into the control system of the unmanned aerial vehicle.
  5. 5. The unmanned ship control method according to claim 4, wherein a real-time monitoring system is established to monitor and evaluate the unmanned ship control process in real time, and the steps further comprise: acquiring state information, environment sensing data and control instructions of the unmanned ship in real time; preprocessing the received data, extracting features and identifying modes of the perceived data, and evaluating the control state and environmental change of the unmanned ship; displaying the position, heading and speed information of the unmanned ship in real time through a monitoring interface; detecting possible abnormal states or potential risks in real time according to control data and environment information of the unmanned ship; the control performance of the unmanned ship is evaluated in real time, and the evaluation result is fed back to an operator or a control system; and establishing a data storage mechanism, and performing persistence storage on all data generated in the monitoring process.
  6. 6. The method of unmanned aerial vehicle steering according to claim 5, wherein, Building a test environment; Testing the basic control function of the unmanned ship; quantitatively evaluating the control performance of the unmanned ship; analyzing the test data to find possible errors or shortages in the algorithms and mechanisms; And according to the test result, parameter adjustment, structure improvement or model update is carried out on the deep learning algorithm, and the trigger condition and the intervention strategy of the automatic intervention mechanism are optimized.

Description

Unmanned ship control method Technical Field The invention relates to the technical field of unmanned boats, in particular to a control method of an unmanned boat. Background With the wider application of unmanned ships, the accuracy and stability of the unmanned ships are urgent to be solved. However, in some cases, manual manipulation is still necessary in the unmanned boat manipulation system currently available on the market, and particularly in complex or special task environments, an operator can select a manual control mode through an interactive panel to directly control the forward, reverse, left turn and right turn of the unmanned boat. However, the lack of an effective automatic intervention mechanism results in that the unmanned ship is easy to be influenced by external environment, and the accuracy and stability of the unmanned ship cannot be ensured. Disclosure of Invention The invention aims to provide a control method of an unmanned ship, which solves the problems that the control of the unmanned ship is easily influenced by external environment and the accuracy and stability of the control cannot be ensured due to the fact that the existing unmanned ship control system lacks an effective automatic intervention mechanism. In order to achieve the above purpose, the invention provides a control method of an unmanned ship, comprising the following steps: upgrading a sensing device of the unmanned ship to obtain environment sensing data; Designing and optimizing a deep learning algorithm to extract features from the perception data and generate a control instruction; based on a deep learning algorithm, designing an automatic intervention mechanism; A real-time monitoring system is established to monitor and evaluate the control process of the unmanned ship in real time; And testing and verifying the unmanned ship, and continuously optimizing and adjusting a deep learning algorithm and an automatic intervention mechanism according to the test result. The unmanned ship sensing device is upgraded to acquire environment sensing data, and the method further comprises the following steps: and introducing a multi-sensor fusion technology, and carrying out fusion processing on data collected by different sensors. Wherein the deep learning algorithm is designed and optimized to extract features from the perceptual data and to generate control instructions, the steps further comprising: collecting and preparing perception data; Selecting a deep learning model architecture according to task requirements and data characteristics; designing a network layer and an activation function, performing feature pre-training by using an unsupervised learning method, introducing an attention mechanism or a feature fusion technology, and enhancing the attention capability of an algorithm to key features; Designing an output layer; Model training is carried out by using the loss function and an optimization algorithm; and evaluating the trained model by using the verification set, analyzing the performance of the model, and optimizing the model according to the evaluation result. Wherein, based on the deep learning algorithm, an automatic intervention mechanism is designed, and the steps further comprise: collecting control data of the unmanned ship in various environments and conditions, marking the data, and determining which conditions need automatic intervention and the targets and expected results of the intervention; training a deep learning model by using the marked data, so that the deep learning model can identify a scene needing automatic intervention and forecast a corresponding control instruction; According to the result of model prediction, designing an automatic intervention strategy; The automatic intervention mechanism is integrated into the control system of the unmanned aerial vehicle. The method comprises the steps of establishing a real-time monitoring system, and carrying out real-time monitoring and evaluation on the control process of the unmanned ship, wherein the steps further comprise: acquiring state information, environment sensing data and control instructions of the unmanned ship in real time; preprocessing the received data, extracting features and identifying modes of the perceived data, and evaluating the control state and environmental change of the unmanned ship; displaying the position, heading and speed information of the unmanned ship in real time through a monitoring interface; detecting possible abnormal states or potential risks in real time according to control data and environment information of the unmanned ship; the control performance of the unmanned ship is evaluated in real time, and the evaluation result is fed back to an operator or a control system; and establishing a data storage mechanism, and performing persistence storage on all data generated in the monitoring process. Wherein, a test environment is built; Testing the basic control function of the unmanned ship; qu