Underwater robot intelligent navigation and environment sensing system based on multi-sensor fusion collaborative technology

Abstract

The invention provides an underwater robot intelligent navigation and environment sensing system based on a multi-sensor fusion collaborative technology. The system integrates a plurality of sensors such as sonar, optical, electromagnetic and the like, and realizes comprehensive data acquisition and real-time perception of an underwater environment. By applying advanced algorithms such as Kalman filtering and a neural network, the system performs synchronization, preprocessing and deep fusion on multi-source data, and the consistency and accuracy of information are remarkably improved. In addition, the system also has adaptive navigation and environment perception capabilities, can dynamically adjust a navigation path and a perception strategy according to fused data, and adapts to a complex and changeable underwater environment. And the intelligent decision support system provides decision assistance for an operator in combination with an environment perception result and a task target. And due to the design of a user interaction and monitoring interface, the usability and real-time monitoring capability of the system are ensured. According to the implementation of the invention, the precision and efficiency of the underwater robot in tasks such as monitoring and exploration are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater robots, and particularly to an intelligent navigation and environment perception system that integrates multiple sensors and adopts advanced data fusion algorithms. Background Art

[0002] In the current technical field of underwater robots, the performance of the navigation and environment perception system is a key factor restricting its application scope and efficiency. Traditional underwater robots mostly use a single sensor for data collection, such as a sonar sensor. Although it can provide distance and azimuth information in some cases, in the face of the complexity of the underwater environment, such as turbid water quality and strong current interference, its performance is limited and it is difficult to meet the requirements of high-precision navigation and comprehensive environment perception. Although optical sensors can provide high-resolution images, they are affected by lighting conditions and visibility; although electromagnetic sensors are sensitive to certain specific applications, they also have limitations in applicable conditions.

[0003] Although there have been attempts to integrate multiple sensors into underwater robots in order to improve the overall performance of the system through data fusion, the existing technologies still face many challenges in practical applications. First, there are differences in time synchronization and spatial registration of the data collected by different sensors, resulting in difficulties in ensuring the consistency and accuracy of the fused data. Second, the multi-sensor data fusion algorithm itself is highly complex and requires a large amount of computing resources, while the computing power of underwater robots is limited, which restricts the real-time performance and reliability of the algorithm. Third, the existing systems are insufficient in terms of adaptive adjustment and intelligent decision support, and it is difficult to cope with the dynamic changes of the underwater environment and complex task requirements.

[0004] It is necessary to develop a new type of underwater robot multi-sensor fusion intelligent navigation and environment perception system, which should have a high-precision data synchronization mechanism, advanced data fusion algorithms, powerful real-time processing capabilities, and intelligent decision support functions. Such a system can make full use of the advantages of each sensor, overcome the deficiencies of single sensors, achieve stable and reliable navigation and comprehensive and accurate environment perception in a variable underwater environment, thereby significantly improving the operation ability and application scope of underwater robots.

[0005] Technical Solution

[0006] The intelligent navigation and environment perception system for underwater robots with multi-sensor fusion proposed by the present invention is realized through the following specific technical solutions:

[0007] The sensor integration unit constructs a highly integrated sensor unit, which integrates the following key sensor components:

[0008] Sonar sensor: Equipped with a transmitter with adjustable frequency, which can emit broadband acoustic signals to penetrate the underwater environment, and uses at least two highly sensitive receivers to receive the reflected acoustic waves. The design of the transmitter and receivers allows the system to adjust the operating frequency under different underwater conditions to obtain the best signal propagation and obstacle detection effects. In addition, by measuring the round-trip time of the acoustic waves, the system can accurately calculate the distance and azimuth of the obstacles, providing key obstacle avoidance information for the underwater robot.

[0009] Optical imaging sensor: Includes one or more high-resolution cameras, which can stably capture images under various underwater lighting conditions. The image processing unit uses advanced image recognition and enhancement algorithms to perform real-time processing on the captured images, realizing the recognition of the shape, size and distance of objects, as well as clear imaging in low-light environments.

[0010] Electromagnetic field sensor: At least includes a pair of electric field and magnetic field detectors, which are highly sensitive and can detect weak electromagnetic field changes in the underwater environment. The electric field detector is used to sense changes in underwater current or artificial electromagnetic sources, while the magnetic field detector is used to identify local changes in the geomagnetic field or magnetic field anomalies caused by metal objects, providing auxiliary navigation information for the underwater robot.

[0011] The design of this sensor integration unit takes into account the operating environment and task requirements of the underwater robot, ensuring the stability, reliability and environmental adaptability of the sensors. Through precise sensor integration and data processing, the underwater robot of the present invention can achieve a comprehensive perception of the underwater environment, providing a solid foundation for intelligent navigation and environmental perception.

[0012] The data acquisition and preprocessing module is carefully designed for each type of sensor in the sensor integration unit of the underwater robot, ensuring high-precision and high-reliability of the data. The following are the specific details of this module:

[0013] Dedicated data acquisition circuit: An independent data acquisition circuit is designed for each sensor, with the ability to precisely match the output characteristics of the sensor, including impedance matching, signal gain adjustment, etc., to ensure the fidelity of the signal during the acquisition process.

[0014] High-precision analog-to-digital converter (ADC): An industry-leading ADC is selected, with a resolution of at least 24 bits and a sampling rate higher than the highest frequency of the sonar sensor, ensuring that subtle changes in the signal can be accurately captured during the digitization process.

[0015] Signal Conditioning: For sonar sensors, a signal conditioning link including a low-noise amplifier (LNA) and a band-pass filter is designed to enhance the signal and filter out noise; for optical imaging sensors, automatic gain control (AGC) and white balance adjustment are implemented to adapt to different lighting conditions.

[0016] Real-time Feature Extraction Algorithm: An efficient real-time algorithm is developed to extract key features from digitized signals. For example, feature extraction of sonar signals includes peak detection to determine the distance to obstacles and frequency analysis to identify the material of obstacles.

[0017] Data Synchronization and Timestamping: Through a high-precision global clock system, a unified time reference is provided for all sensor data streams to ensure data synchronization, and an accurate timestamp is attached to each data packet.

[0018] Data Optimization and Denoising: Advanced denoising techniques such as wavelet transform or adaptive filters are applied to reduce random noise in the signal and improve data quality.

[0019] Data Formatting and Encoding: The optimized data is converted into a unified format and necessary encoding is performed to reduce data redundancy and improve storage and transmission efficiency.

[0020] Modular and Scalable Design: The preprocessing module adopts a modular design, which is convenient for adding or replacing specific signal processing algorithms according to different task requirements and adapting to the development of future sensor technologies.

[0021] Real-time Performance and Resource Optimization: All algorithms and processing flows are optimized to meet the requirements of real-time processing, and considering the computing resource limitations of the underwater robot, the response speed and processing ability of the system are ensured.

[0022] Through these specific and refined designs, the data acquisition and preprocessing module of the present invention provides a solid, reliable and efficient data foundation for the underwater robot, and provides strong support for subsequent intelligent navigation and environmental perception tasks.

[0023] The real-time data synchronization mechanism, a key technical link to ensure the accuracy of multi-sensor data fusion, the specific details are as follows:

[0024] High-precision Clock Source: The system is equipped with a high-precision clock source, such as a cesium atomic clock or a high-precision quartz crystal oscillator, to provide a unified time reference for the entire underwater robot.

[0025] Global Time Synchronization Protocol: A global time synchronization protocol such as NTP or PTP is adopted to ensure that the time of all sensors and processing units in the system is synchronized to the microsecond level or even higher precision.

[0026] Timestamp technology: The data acquisition unit of each sensor embeds an accurate timestamp in the data packet to record the real time when the data is generated, ensuring that the data can be accurately aligned during processing.

[0027] Synchronous signal distribution: Through a dedicated synchronous signal distribution network, the clock signal or synchronous pulse is distributed to the data acquisition circuit of each sensor to achieve synchronous triggering of data acquisition.

[0028] Precise distribution of timestamps: Using FPGA (Field Programmable Gate Array) or dedicated synchronous logic circuits, timestamps are precisely distributed while data is being acquired, reducing synchronization errors.

[0029] Data buffering and alignment: Before the data is transmitted to the central processing unit, a buffering mechanism is used to temporarily store the data, and the data is sorted and aligned according to the timestamps to ensure the temporal consistency of the data.

[0030] Synchronization error correction: A synchronization error correction algorithm has been developed to identify and compensate for time deviations caused by signal propagation delays or processing delays.

[0031] Dynamic synchronization adjustment: The system can dynamically adjust the synchronization strategy according to the synchronism errors monitored in real time to adapt to changes in the underwater environment or the working state of the sensors.

[0032] Synchronization status monitoring: The synchronization status is monitored in real time, and synchronization events and errors are recorded through the system log to provide data support for system maintenance and performance optimization.

[0033] Redundant synchronization mechanism: A redundant synchronization mechanism has been designed. When the primary synchronization link has problems, it can automatically switch to the backup synchronization link to ensure the stable operation of the system.

[0034] Through these specific technical measures, the real-time data synchronization mechanism of the present invention ensures the strict temporal consistency of the data streams from different sensors, provides a solid temporal benchmark for multi-sensor data fusion, and thus improves the accuracy and reliability of the intelligent navigation and environmental perception system of the underwater robot.

[0035] Regarding the multi-sensor data fusion algorithm, an advanced multi-sensor data fusion algorithm has been developed. This algorithm realizes the real-time fusion processing of multi-source heterogeneous sensor data by combining improved Kalman filtering technology and deep learning technology. The following are the specific implementation details of this algorithm:

[0036] Algorithm framework: A multi-level data fusion framework has been constructed. First, local data preprocessing is performed at the sensor level, then common information is extracted at the feature level, and finally all information is integrated at the decision level to generate a unified environmental model.

[0037] Improved Kalman Filtering: Improved methods such as Extended Kalman Filter (EKF) or Unscented Kalman Filter (UKF) are adopted to handle the filtering problems of nonlinear systems, so as to more accurately estimate the state of the underwater robot.

[0038] Deep Learning Integration: Deep learning models such as Convolutional Neural Network (CNN) or Recurrent Neural Network (RNN) are utilized to perform feature learning and pattern recognition on sensor data, in order to improve the accuracy and robustness of data fusion.

[0039] Real-time Data Processing: The algorithm is designed in a stream processing mode, capable of processing real-time data streams, quickly updating the environmental model, and meeting the underwater robot's requirement for rapid response.

[0040] Heterogeneous Data Fusion: A dedicated fusion strategy is developed to process different types of data, such as fusing the distance measurement data of sonar, the image data of optical imaging, and the field strength data of electromagnetic fields into a unified environmental description.

[0041] Uncertainty Quantification: The uncertainty of the environmental model is evaluated through the fusion algorithm, such as estimating the error covariance, to provide a reliability index for decision-making.

[0042] Adaptive Filter Gain Adjustment: According to environmental changes and the confidence of sensor data, the filter gain is dynamically adjusted to optimize the data fusion performance.

[0043] Multi-model Fusion Strategy: When dealing with sensor data with different characteristics, fusion strategies such as weighted average, Bayesian method, or the output of neural networks are adopted to achieve optimal information integration.

[0044] Optimization of the Fusion Algorithm: The algorithm has been computationally optimized to reduce the computational complexity and ensure that the algorithm can operate efficiently on the limited computational resources of the underwater robot.

[0045] Algorithm Verification and Testing: Through tests in simulated and actual underwater environments, the effectiveness and accuracy of the fusion algorithm have been verified.

[0046] Through these specific implementations, the multi-sensor data fusion algorithm of the present invention can generate an accurate and real-time updated environmental model, providing strong information support for the intelligent navigation and environmental perception of the underwater robot.

[0047] The described adaptive parameter adjustment strategy utilizes the real-time feedback and prediction mechanism of the environmental model to intelligently adjust the key parameters in the data fusion algorithm. This strategy is implemented through a closed-loop control system, where environmental perception data is continuously monitored and compared with preset performance indicators. Based on this data, the adjustment unit employs advanced optimization algorithms, such as genetic algorithms or particle swarm optimization, to dynamically calculate and update the filter gain to adapt to the impact of environmental changes on sensor readings. At the same time, the unit also reallocates sensor weights to ensure that in a variable underwater environment, the system can rely on the most reliable and accurate sensor inputs. For example, in low visibility conditions, the weight of the sonar sensor may be increased, while in good lighting conditions, the weight of the optical imaging sensor may be increased. In addition, the adaptive algorithm adjustment unit also includes a prediction component that uses time series analysis or machine learning models to predict environmental change trends and pre-adjust parameters to optimize the future data fusion process. This adaptive adjustment not only improves the system's response speed and accuracy but also enhances the stability and reliability of the underwater robot in complex environments.

[0048] The described intelligent navigation decision-making system integrates advanced path planning and cost assessment functions to ensure the efficient and safe operation of the underwater robot. The following is the detailed sequence and characteristics of the system operation:

[0049] Initial environmental perception: The system first receives and processes the environmental model data from the multi-sensor fusion algorithm, which forms the basis for path planning.

[0050] Path planning algorithm initialization: Using the A* or D* algorithm, the path planning process is initialized according to the current environmental model data to determine the starting point and the target point.

[0051] Dynamic cost assessment: The system calculates the path cost from the starting point to the target point in real time through a dynamic cost assessment function, considering safety, energy consumption, and time efficiency.

[0052] Path search and optimization: The navigation decision-making system performs path search, and the optimization algorithm selects the best path according to the cost assessment results, considering avoiding known obstacles and potential risk areas.

[0053] Real-time environmental monitoring: During the path execution process, the system continuously monitors environmental changes and updates the environmental model using the multi-sensor data fusion algorithm.

[0054] Path update and replanning: In case of environmental changes or the emergence of new obstacles, the system will update the path in real time or trigger replanning to ensure the safety of the robot and the continuity of the task.

[0055] Prediction and risk management: The system uses prediction algorithms to evaluate environmental trends and potential risks and adjusts the path to avoid predicted risk points.

[0056] User Interaction and Target Adjustment: The operator can input new targets or adjust task parameters through the user interface, and the system takes these changes into account and adjusts the path accordingly.

[0057] Implementation of Multi-Mode Navigation Strategy: The system selects appropriate navigation modes, such as exploration, tracking, or emergency avoidance, according to the current task requirements and environmental conditions.

[0058] Resource Optimization and Computational Efficiency: Ensure that the navigation decision-making algorithm runs efficiently with limited computational resources, and optimize the computational process to reduce latency.

[0059] Feedback and Decision Support: The system provides the path planning results, predicted arrival time, risk assessment, and alternative solutions through the user interface to provide decision support for the operator.

[0060] Modularity and Scalability: The software architecture of the intelligent navigation decision-making system adopts a modular design, which is convenient for future upgrades and function expansions.

[0061] According to this operating sequence, the intelligent navigation decision-making system of the present invention can provide continuous and real-time navigation decision support for the underwater robot, ensuring the efficient and safe execution of the task.

[0062] The obstacle avoidance and behavior adjustment unit is a highly integrated intelligent system specially designed for real-time processing of obstacle avoidance tasks. This unit first uses advanced real-time obstacle detection algorithms to quickly identify obstacles in the underwater environment and classifies the obstacles through a deep learning classifier to distinguish their nature and possible impacts. Once an obstacle is detected, the obstacle avoidance strategy generator immediately works to calculate the optimal obstacle avoidance path, taking into account the current speed, direction, and dynamic characteristics of the robot. Using an accurate feedback control system, this unit dynamically adjusts the navigation path and behavior parameters of the robot, such as speed and heading, to ensure a smooth transition and safe distance during obstacle avoidance. In addition, this unit also has learning and adaptation capabilities, and can optimize the obstacle avoidance strategy according to historical obstacle avoidance data to improve the efficiency and accuracy of obstacle avoidance. Through this integrated obstacle avoidance solution, the underwater robot can effectively cope with various sudden obstacles while ensuring task continuity, significantly improving its autonomy and safety in complex underwater environments.

[0063] The user interaction and monitoring interface is designed to comprehensively present the sensor data, environmental model, navigation path, and system status information of the underwater robot to the operator in an intuitive graphical display manner. Utilizing the latest graphical user interface (GUI) technology, this interface not only enables real-time visualization of data but also provides detailed system alerts and operation logging functions. The operator can monitor the real-time position, path changes, sensor readings, and environmental characteristics of the robot through this interface, ensuring a comprehensive understanding of the robot's status. In addition, the interface integrates manual intervention tools, allowing the operator to input commands through an intuitive control panel when necessary, adjust sensor parameters, or directly control the navigation behavior of the robot. To enhance the user experience, the interface design emphasizes ease of use and response speed, ensuring quick responses in emergency situations. Meanwhile, help documents and operation guides are provided to reduce the operation difficulty and improve the accessibility of the system. Through this highly integrated and user-friendly monitoring interface, the present invention significantly improves the convenience and flexibility of underwater robot operation.

[0064] The operation control panel is a highly integrated and intuitive interface that provides the operator with comprehensive manual control capabilities. The panel is equipped with easily recognizable buttons and controls, allowing the user to perform various critical operations, such as sending an emergency stop command to quickly respond to potential hazards, modifying the navigation path of the robot in real time to adapt to changes in task requirements, and switching sensor modes according to environmental conditions to optimize data collection. The control panel also has the function of adjusting sensor parameters, enabling the user to fine-tune the sensor settings according to specific tasks or environmental characteristics to obtain optimal performance. In addition, status indicators and displays are provided on the panel to give the user real-time feedback, including system status, alert information, and operation result confirmation. Through the carefully designed layout and user-friendly operation logic, the control panel ensures that the operator can quickly and accurately perform manual intervention even under pressure or in emergency situations, thereby enhancing the flexibility and operation safety of the underwater robot.

[0065] The system integration and testing phase of the present invention is a comprehensive and rigorous process aimed at ensuring that all modules and units of the underwater robot can work efficiently in cooperation and meet the performance standards as designed. In this phase, first, individual functional tests are conducted on each module to verify its independent performance in a simulated underwater environment. Subsequently, all modules are integrated into the underwater robot platform for system-level integration testing to ensure seamless connection and interoperability among data streams, control signals, and user interfaces. The test content covers the entire system process from sensor data acquisition, real-time data processing, multi-sensor data fusion, intelligent navigation decision-making, obstacle avoidance and behavior adjustment to user interaction and monitoring interfaces. After the system integration testing, the system will be subjected to preliminary on-site testing in a controlled pool or simulated underwater environment to evaluate its performance under actual underwater operation conditions. This includes the evaluation of sensor accuracy, navigation accuracy, obstacle avoidance efficiency, and user interface responsiveness. Then, the system will be subjected to in-depth on-site testing in open water to further verify its stability and reliability in a real ocean environment, including its performance under dynamic ocean currents, changing seabed topography, and different lighting conditions. The system integration and testing phase also includes the evaluation of system security and redundancy to ensure that in the event of a failure of a critical component, the system can operate safely in a degraded mode or automatically switch to a standby mode. Finally, through this series of comprehensive tests, the underwater robot platform of the present invention will demonstrate the powerful performance of its highly integrated system and its wide application potential in diverse underwater tasks.

[0066] The following are the detailed operation steps for the underwater robot system:

[0067] Step 1, initialize and calibrate all sensors: When starting the underwater robot, the system automatically runs a self-check program to conduct functional tests and performance calibrations on all sensors. The calibration process includes the transmission frequency and receiving sensitivity of sonar sensors, the exposure and white balance of optical imaging sensors, and the signal resolution of electromagnetic field sensors. The calibration parameters are optimized according to the specifications of the sensors and historical calibration data to ensure the accuracy and consistency of the collected data.

[0068] Step 2, synchronize sensor data through the central processing unit and perform preprocessing: The central processing unit receives data streams from each sensor, synchronizes the data timestamps using a high-precision clock to ensure data synchronization. Preprocessing is performed on the synchronized data, including using a band-pass filter to remove high-frequency noise and low-frequency drift from sonar signals, applying adaptive gain control and dynamic range compression to optical image data. Data normalization is carried out to convert data from different sensors into a unified numerical range for subsequent processing and fusion.

[0069] Step 3: Use a data fusion processor to fuse multi-source data: The data fusion processor adopts an improved Kalman filtering algorithm and combines it with a deep learning model to fuse the preprocessed data. The fusion process generates a comprehensive environment model, including the positions of obstacles, the contours of underwater terrain, the distribution of electromagnetic fields, etc. The uncertainty quantification unit evaluates the accuracy and reliability of the environment model and provides a confidence score and an error range.

[0070] Step 4: The navigation decision-making unit calculates the optimal path: The navigation decision-making unit receives the comprehensive environment model and task parameters and uses a heuristic search algorithm to calculate the optimal path from the current position to the target position. The path planning takes into account safety, energy consumption, and time efficiency, avoids high-risk areas, and conforms to the task requirements set by the user.

[0071] Step 5: The obstacle avoidance unit monitors in real time and generates an obstacle avoidance strategy: The obstacle avoidance unit receives sensor data in real time and uses a machine learning classifier to quickly identify obstacles. According to the characteristics of the obstacles and the dynamic model of the robot, it generates an obstacle avoidance strategy, such as detouring, pausing, or speed adjustment. The navigation decision-making unit dynamically adjusts the optimal path according to the obstacle avoidance strategy to ensure the smoothness and efficiency of the obstacle avoidance action.

[0072] Step 6: The user monitors the system status through the interaction interface: The user interaction interface provides real-time data visualization, including the environment model, path planning, and sensor status. The user can manually adjust the sensor parameters according to the monitoring information, such as the operating frequency of the sonar sensor or the resolution of the optical sensor. In case of an emergency, the user can manually intervene in the robot's behavior through the control panel, such as sending an emergency stop command or adjusting the heading.

[0073] Step 7: The system executes the optimal path and automatically activates the obstacle avoidance strategy: The system controls the underwater robot to move along the path according to the calculated optimal path and the dynamically updated environment model. When encountering unknown obstacles, the obstacle avoidance unit is automatically activated to adjust the robot's behavior in real time to avoid collisions. The system continuously monitors the environment and the robot's status during the entire task execution to ensure the smooth completion of the task.

[0074] Through these detailed operation steps, the underwater robot system of the present invention can achieve efficient and accurate data collection, real-time and accurate environment perception, as well as safe and reliable autonomous navigation and obstacle avoidance.

[0075] Adopting the above technical solutions can bring the following three specific and profound technical effects:

[0076] 1. Enhanced multi-dimensional environmental perception accuracy: By integrating high-frequency sonar sensors, the system can emit and receive broadband sound waves. Using the time delay and Doppler effect of sound waves, it can accurately measure the distance, azimuth, and speed of surrounding obstacles, and maintain high-resolution obstacle detection even in extremely low visibility environments. The optical imaging sensor, equipped with a high-resolution camera and advanced image processing algorithms, can achieve clear recognition and classification of objects under various lighting conditions, providing rich visual information such as color, texture, and shape. The high-sensitivity detector of the electromagnetic field sensor can capture weak electromagnetic changes in the underwater environment, providing indirect information about the presence of artificial structures such as underwater cables and pipelines for underwater robots.

[0077] 2. Optimized dynamic path planning and autonomous navigation performance: The intelligent navigation decision-making system uses advanced A* or D* path planning algorithms, combined with a cost evaluation function, considering the safety, energy consumption, and time efficiency of the path, to calculate the optimal navigation path, ensuring that the underwater robot completes tasks in the most economical and safe way. The adaptive parameter adjustment strategy dynamically adjusts the parameters of the fusion algorithm, such as filter gains and sensor weights, by real-time analyzing the environmental model and predicting environmental changes, enabling the system to quickly adapt to changes in environmental factors such as water flow and temperature, and improving the response speed and accuracy of autonomous navigation.

[0078] 3. Efficient real-time obstacle avoidance and emergency response mechanism: The obstacle avoidance unit integrates advanced real-time obstacle detection and classification algorithms, which can quickly identify sudden obstacles during the movement of the underwater robot and generate obstacle avoidance strategies. Through a precise control system, it adjusts the heading and speed of the robot to achieve agile obstacle avoidance. The user interface provides a detailed graphical display, including real-time sensor data streams, environmental models, and obstacle avoidance action feedback, enabling users to intuitively monitor the operation status of the underwater robot. In case of emergency, the operator can quickly intervene through the operation control panel to perform operations such as emergency stop or path adjustment to ensure the safety of the task and the reliability of the robot.

[0079] Through these technical effects, the underwater robot can achieve more accurate perception, more intelligent decision-making, and more flexible response in complex and dynamically changing underwater environments, greatly enhancing its practicality and efficiency in application scenarios such as ocean exploration, underwater construction monitoring, and rescue operations. Brief Description of the Drawings

[0080] Figure 1 It is a schematic diagram of the system module of the present invention

[0081] Figure 2 It is a flowchart of the system operation of the present invention Detailed Embodiment

[0082] Next, it will be combined with the appendix in the embodiments of the present inventionFigure 1 and Figure 2 The technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0083] An underwater robot intelligent navigation and environmental perception system based on multi-sensor fusion and collaboration technology, and the specific implementation is as follows:

[0084] Sensor initialization and calibration: When the system starts, first initialize the sonar sensor, set the transmission frequency to 5 MHz, and calibrate the gains of at least two receivers to ensure the consistency of signal reception. Initialize the optical imaging sensor, set the camera resolution to 1920×1080 pixels, and the autofocus range to 0.5 meters to infinity. Initialize the electromagnetic field sensor, set the sensitivity of the electric field detector to 0.05 mV / m, and the sensitivity of the magnetic field detector to 0.001 nT.

[0085] Data acquisition and preprocessing: The sonar sensor emits sound waves, and the receivers synchronously receive the echoes. The data acquisition frequency is 10 kHz and is converted by a 16-bit ADC. The optical imaging sensor captures images at a rate of 30 frames per second, and the image processing unit executes edge detection and object recognition algorithms in real time. The electromagnetic field sensor acquires data at a frequency of 1 kHz, is converted by a 12-bit ADC, and applies an anti-interference filter.

[0086] Data synchronization: Use a high-precision clock source synchronized by GPS to synchronize all sensor data to an accuracy within 1 microsecond. The timestamp embedded in the data packet has an accuracy of 100 nanoseconds to ensure the accurate alignment of data in the central processing unit.

[0087] Multi-sensor data fusion: The data fusion processor adopts the unscented Kalman filter algorithm, with a fusion period of 50 milliseconds and a processing delay not exceeding 30 milliseconds. The update frequency of the generated environmental model is 5 Hz, and the model accuracy error is controlled within 3%.

[0088] Adaptive parameter adjustment: The adaptive algorithm adjustment unit automatically adjusts the parameters of the Kalman filter once every 10 seconds according to the recognition result of the underwater environmental characteristics. The adjusted parameters include sensor weights, process noise covariance, and observation noise covariance.

[0089] Path planning and cost evaluation: The navigation decision unit uses the A* algorithm to calculate the path, considering path length, estimated passing time, energy consumption, and safety buffer. The cost evaluation function is updated every 5 seconds, and the evaluation result is used to optimize the path planning.

[0090] Obstacle Avoidance Strategy Generation: The obstacle avoidance unit uses a deep learning-based classifier to classify obstacles, with a classification accuracy exceeding 90%. Based on the type and size of the obstacles, the obstacle avoidance strategy generator calculates the obstacle avoidance path within 200 milliseconds.

[0091] User Interaction and Monitoring: The user interface displays sensor data and the environmental model at a refresh rate of 15 frames per second. The operation control panel provides 12 programmable buttons for manually controlling sensor parameters and robot behavior.

[0092] Execution and Feedback Control: The system executes the optimal path with a feedback control cycle of 100 milliseconds, ensuring the robot runs stably along the predetermined path. The execution delay of the obstacle avoidance strategy does not exceed 150 milliseconds, ensuring timely response to sudden obstacles.

[0093] System Integration and Testing: The system is integrated and tested in a simulated underwater environment. The tests include sensor accuracy tests, data synchronization tests, fusion algorithm tests, and navigation decision tests. Field tests include static tests in a controlled pool and dynamic tests in open water to verify the performance of the system in an actual underwater environment.

[0094] Technical Parameters and Results:

[0095] Sonar Sensor: Transmission frequency 5 MHz, receiving sensitivity -205 dB, obstacle detection range 0.5 - 50 meters.

[0096] Optical Imaging Sensor: Resolution 1920×1080 pixels, light adaptation range 0.001 - 100,000 lux, object recognition accuracy 95%.

[0097] Electromagnetic Field Sensor: Electric field detector sensitivity 0.05 mV / m, magnetic field detector sensitivity 0.01 nT, detection range 0.1 - 10 nT.

[0098] Data Fusion Processor: Fusion cycle 50 milliseconds, processing delay 30 milliseconds, environmental model update frequency 5 Hz, accuracy error <3%.

[0099] Navigation Decision Unit: Path planning time <1 second, optimization considerations include path length, time, energy consumption, and safety buffer.

[0100] Obstacle Avoidance Unit: Obstacle detection response time <0.5 second, obstacle avoidance strategy generation time <0.2 second, obstacle avoidance action execution delay <0.15 second.

[0101] Through these detailed implementation steps and parameter settings, the underwater robot system of the present invention can achieve precise environmental perception, intelligent path planning, effective obstacle avoidance response, and a user-friendly interaction experience, meeting the requirements of complex underwater operation tasks.

[0102] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and the descriptions in the specification are only preferred examples of the present invention, and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. An underwater robot intelligent navigation and environment perception system based on multi-sensor fusion and collaborative technology, characterized in that: The system includes: Sonar sensors have the function of transmitting broadband sound waves and receiving echoes through at least two receivers to determine the distance and direction of obstacles; Optical imaging sensors, including high-resolution cameras and image processing units, for capturing underwater images and performing object recognition and classification; An electromagnetic field sensor, comprising at least an electric field detector and a magnetic field detector, for detecting changes in the electromagnetic field in an underwater environment; The central processing unit integrates the data fusion processor, adaptive algorithm adjustment unit, navigation decision unit, obstacle avoidance unit, user interaction interface and operation control panel; Data fusion processor, using improved Kalman filter algorithm, is used to fuse multi-sensor data in real time to provide accurate environmental model; Adaptive algorithm adjustment unit dynamically adjusts the parameters of the fusion algorithm according to environmental changes to optimize navigation accuracy; The navigation decision unit uses the A* or D* algorithm to calculate the optimal path and uses the cost evaluation function to evaluate the safety, energy consumption and time efficiency of the path; The obstacle avoidance unit monitors obstacles on the path in real time and generates obstacle avoidance strategies to adjust the robot’s navigation path; User interface that graphically displays sensor data, environmental models, navigation paths, and system status, including a status indicator module that displays system status and alarm information; The operation control panel allows the user to manually input commands, adjust sensor parameters and intervene in robot behavior.

2. The system according to claim 1, characterized in that The transmitter of the sonar sensor has a function of adjusting the transmission frequency to adapt to different underwater environmental conditions.

3. The system according to claim 1, characterized in that The image processing unit of the optical imaging sensor comprises: Object recognition algorithms, used to identify specific objects in images in real time; Shape estimation module, which is used to measure the size and shape of objects in images.

4. The system according to claim 1, characterized in that The electric field detector and the magnetic field detector of the electromagnetic field sensor have high sensitivity and can accurately detect changes in weak electromagnetic fields.

5. The system according to claim 1, characterized in that The data fusion processor of the central processing unit further comprises: A state estimation unit, used for estimating the six-degree-of-freedom motion state of the underwater robot; Uncertainty quantification unit to evaluate the accuracy and reliability of state estimates.

6. The system according to claim 1, characterized in that The adaptive algorithm adjustment unit comprises: Environmental feature recognition module, used to analyze underwater environmental features and identify environmental changes; Parameter optimization algorithm is used to automatically adjust the parameters of the Kalman filter based on the environmental feature recognition results.

7. The system according to claim 1, characterized in that The path planning algorithm of the navigation decision unit includes: Heuristic search strategy for quickly finding feasible paths in complex underwater environments; Multi-objective optimization algorithm is used to consider multiple performance indicators simultaneously when planning the path.

8. The system according to claim 1, wherein the obstacle detection module of the obstacle avoidance unit comprises: Real-time data processing capability to quickly respond to sudden obstacles; Machine learning classifiers to improve the accuracy of obstacle detection.

9. The system according to claim 1, characterized in that The user interaction interface provides: multi-mode display options, allowing the user to select different display modes according to their needs; Interactive help system provides users with operation guidance and system status explanation.

10. An operating method using the system according to any one of claims 1 to 9, comprising the following steps: Step 1: Initialize and calibrate all sensors to ensure the accuracy of data collection; Step 2: Synchronize sensor data through the central processing unit and perform preprocessing, including filtering and normalization; Step 3: Use the data fusion processor to fuse multi-source data to generate a comprehensive environmental model, and the uncertainty quantification unit evaluates the reliability of the model; Step 4: The navigation decision unit uses the path planning algorithm to calculate the optimal path based on the comprehensive environment model and task requirements; Step 5: The obstacle avoidance unit monitors obstacles in real time and generates an obstacle avoidance strategy, and the navigation decision unit adjusts the optimal path according to the obstacle avoidance strategy; Step 6: The user monitors the system status through the interactive interface and manually adjusts sensor parameters or intervenes in robot behavior as needed; In step seven, the system executes the calculated optimal path and automatically initiates obstacle avoidance strategy when encountering unknown obstacles to ensure the successful completion of the task.

Images