Unmanned ship navigation system based on vision and biological monitoring

Abstract

The invention provides an unmanned ship navigation system based on vision and biological monitoring, and aims to solve the problems of insufficient navigation precision and safety of an unmanned ship in a complex water area. The system comprises a visual perception module, a biological monitoring module, a navigation control module and a communication module. The visual perception module acquires environment information by using a high-resolution camera and a laser radar, and identifies obstacles through a multi-scale convolutional neural network; the organism monitoring module adopts an underwater sonar system and a Bayesian classification algorithm to detect and identify underwater organisms; the navigation control module combines an A * algorithm and a dynamic time warping algorithm to carry out path planning and real-time adjustment; and the communication module realizes secure data transmission through a symmetric encryption algorithm. The system can navigate autonomously in a complex dynamic environment, avoids obstacles and organisms, ensures navigation safety, causes minimum interference to an underwater ecological environment, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned ship navigation systems, and in particular to an unmanned ship navigation system based on vision and biological monitoring. Background Art

[0002] With the development of unmanned ship technology, its application in military, scientific research, environmental monitoring, logistics and transportation is becoming more and more extensive. Unmanned ships have gradually become an indispensable tool in marine and inland river environments due to their advantages such as high efficiency, precision and unmanned operation. However, in complex and changing water environments, traditional unmanned ship navigation systems still face many challenges, which are mainly concentrated in the following aspects:

[0003] Environmental perception and obstacle recognition: Traditional unmanned ship navigation systems mainly rely on GPS and inertial navigation systems for path planning. However, in environments with many obstructions (such as ports, estuaries, etc.) or areas with poor signals (such as canyons, bays, etc.), GPS signals are often interfered with, resulting in reduced navigation accuracy. In addition, traditional navigation systems have limited recognition capabilities when facing dynamic obstacles such as floating objects, fishing nets, and buoys on the water surface, making it difficult to effectively avoid obstacles. In this case, the unmanned ship may collide or deviate from the scheduled route, increasing operational risks.

[0004] Monitoring and avoidance of underwater life: Unmanned ships may have an impact on underwater life during navigation. For example, the noise generated by underwater sonar or propulsion systems may interfere with the behavior of marine life and may even cause injury or death to marine life. In addition, existing unmanned ship navigation systems generally lack the ability to detect and avoid underwater life, making it difficult to navigate safely without destroying the underwater ecological environment. This ecological impact is particularly evident in tasks such as marine scientific research and environmental monitoring.

[0005] In complex waters, such as rocky coastlines, narrow rivers, and ports with many ships, the difficulty of path planning increases significantly. The path planning algorithms of traditional navigation systems are often inefficient when dealing with multiple targets and constraints, and it is difficult to generate the optimal route in real time. At the same time, the ability to adjust the route in real time in a dynamic environment is relatively insufficient, which can easily cause unmanned ships to enter dangerous areas.

[0006] When unmanned ships are performing their missions, they need to transmit a large amount of navigation data, environmental information, and mission status back to the remote control center. Traditional communication systems have slow data transmission speeds and are susceptible to interference over long distances and in complex environments. In addition, data can be easily eavesdropped or tampered with during transmission, posing certain security risks. This is a significant problem for missions that require confidentiality or the transmission of sensitive data.

[0007] Based on the above technical challenges, there is an urgent need for an unmanned ship navigation system that can efficiently navigate in complex waters and has the ability to perceive the environment, identify obstacles, and monitor underwater organisms. The system should be able to plan and adjust routes in real time in complex dynamic environments to ensure that the unmanned ship can complete its mission safely and reliably under various harsh conditions. In addition, the system should also have a safe and efficient data transmission mechanism to ensure the security and integrity of mission information.

[0008] The present invention proposes an unmanned ship navigation system based on vision and biological monitoring, aiming to solve the above problems faced by traditional unmanned ship navigation systems. By introducing advanced visual perception technology, biological monitoring technology and intelligent path planning algorithms, the unmanned ship navigation system of the present invention can achieve high-precision, low-risk autonomous navigation in complex waters, and effectively protect the underwater ecological environment, with strong adaptability and broad application prospects. Summary of the invention

[0009] The purpose of the present invention is to provide an unmanned ship navigation system based on vision and biological monitoring to solve the problems raised by the above background technology.

[0010] To achieve the above object, the present invention provides the following technical solution: an unmanned ship navigation system based on vision and biological monitoring, the system comprising:

[0011] A visual perception module, used to obtain visual information of the surrounding environment of the unmanned ship, the visual perception module includes a high-resolution camera and a laser radar, wherein the camera is used to capture image information of the surrounding environment, and the laser radar is used to measure the distance of the target object and its relative position;

[0012] A biological monitoring module, used to detect the presence of underwater organisms, the biological monitoring module includes an underwater sonar system and a biological recognition algorithm, the underwater sonar system detects the size, location and type of underwater organisms by emitting sound waves and receiving reflected signals;

[0013] A navigation control module, configured to receive data from the visual perception module and the biological monitoring module, wherein the navigation control module generates an optimal route through an adaptive navigation algorithm, wherein the algorithm performs path planning based on fused environmental and biological monitoring data;

[0014] The communication module is used to transmit the location information, environmental data and biological monitoring data of the unmanned ship to the remote control center in real time. The communication module also supports receiving instructions issued by the remote control center.

[0015] As a preferred technical solution of the present invention, the visual information obtained by the visual perception module is used to identify obstacles through the following formula:

[0016]

[0017] Among them, f(x,y) represents the characteristic function of the image, g i (x, y) represents image features extracted from different scales, w i is the weight of each feature, and n is the number of features. This feature function extracts image features at different scales through a convolutional neural network (CNN), and combines the weighted average method to generate image feature representations for obstacle recognition.

[0018] As a preferred technical solution of the present invention, the biometric recognition algorithm is based on a Bayesian classifier and is used to calculate the classification probability of a biological being. The calculation formula is as follows:

[0019]

[0020] Among them, P(c|X) represents the posterior probability of biological category c given feature x, P(X|c) represents the likelihood of feature x under category c, P(c) represents the prior probability of biological category c, and P(X) represents the marginal probability of feature X. The algorithm uses the training data set to model different biological categories and classifies the detected organisms through the principle of maximum a posteriori probability.

[0021] As a preferred technical solution of the present invention, the navigation control module calculates the optimal route by the following formula:

[0022]

[0023] Where L represents the total path length of the route, x i ,y i is the coordinate of the i-th node on the path, m is the number of path nodes, λ is the risk weight coefficient, and R is the risk value detected along the way. The algorithm combines the A* algorithm with the optimization algorithm based on neural networks to calculate the optimal navigation path and can dynamically adjust the path to avoid risks.

[0024] As a preferred technical solution of the present invention, the risk value R is calculated according to the following formula:

[0025]

[0026] Among them, r j represents the risk measure of the jth risk source detected, d j is the distance between the unmanned ship and the jth risk source, and k is the number of risk sources. This formula combines distance and risk measurement to ensure that the unmanned ship can effectively avoid targets that are close and have higher risks.

[0027] As a preferred technical solution of the present invention, the visual perception module recognizes obstacles through a multi-scale convolutional neural network (MSCNN), and the output of the network is multi-layered feature fused through the following formula:

[0028]

[0029] Among them, O represents the output feature after fusion, O l represents the output of the lth layer convolution, α l is the weight coefficient of the lth layer, and L is the number of convolutional layers. This feature fusion technology can weightedly combine image features at different levels to generate more accurate obstacle recognition results.

[0030] As a preferred technical solution of the present invention, the biological monitoring module identifies the organism through a classification model based on the K-nearest neighbor (K-NN) algorithm, and its distance measurement formula is:

[0031]

[0032] Where D(x,y) represents the Euclidean distance between the biological feature vector x and the sample vector y, and p is the feature dimension. The algorithm calculates the distance between biological features and identifies the most similar biological species, thereby achieving accurate identification of aquatic organisms.

[0033] As a preferred technical solution of the present invention, the navigation control module combines the dynamic time warping (DTW) algorithm to match the actual route of the unmanned ship with the planned route, and the calculation formula is:

[0034] DTW(i,j)=min(DTW(i-1,j-1),DTW(i-1,j),DTW(i,j-1))+d(i,j)

[0035] Among them, DTW(i,j) represents the cumulative distance between the i-th point and the j-th point, and d(i,j) represents the local distance between the i-th point and the j-th point. The algorithm can effectively match the actual route with the planned route, evaluate the navigation deviation, and make path corrections when necessary.

[0036] As a preferred technical solution of the present invention, the system includes an adaptive adjustment module, which adjusts the route of the unmanned ship in real time according to environmental changes and biological behaviors, and the adjustment is based on the following formula:

[0037]

[0038] Among them, Δθ represents the adjustment amount of the heading angle, γ is the adjustment coefficient, is the partial derivative of the risk value with respect to the heading angle. This adjustment mechanism can dynamically adjust the navigation path according to the real-time monitored environment and biological information to ensure that the unmanned ship can reach the target location safely and effectively.

[0039] As a preferred technical solution of the present invention, the communication module adopts a symmetric encryption algorithm for data transmission, and the encryption process follows the following formula:

[0040]

[0041] Among them, C is the ciphertext, E(K,M) is the encryption function, M is the plaintext, K is the key, and H(K) is the pseudo-random number generated by the hash function. This encryption mechanism ensures the security of data transmission between the unmanned ship and the remote control center, preventing the data from being eavesdropped or tampered with.

[0042] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0043] Improve navigation accuracy and safety: This invention significantly improves the obstacle recognition and path planning capabilities of unmanned ships in complex water environments by introducing visual perception and biological monitoring modules, combined with multi-scale convolutional neural networks (MSCNN) and Bayesian classification algorithms. The system can accurately identify potential obstacles on the water surface and underwater, and adjust the route in real time to ensure that the unmanned ship can complete the navigation task safely and efficiently.

[0044] Ecological environment protection: Through the introduction of biological monitoring modules, the present invention can monitor and identify underwater biological activities in real time to avoid interference or harm to aquatic organisms. The system fully considers the distribution of underwater organisms when planning the path, and automatically adjusts the route to avoid biological groups, thereby protecting the underwater ecological environment to the greatest extent while performing the task.

[0045] Enhance the adaptability and security of the system: The present invention adopts an adaptive adjustment module and a dynamic time warping (DTW) algorithm, which can quickly adjust the navigation path of the unmanned ship in the face of emergencies or complex environmental changes. At the same time, the communication module uses a symmetric encryption algorithm to ensure the security of data transmission, prevent data leakage or tampering, and further enhance the reliability and adaptability of the unmanned ship system in complex tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a schematic diagram of the structure of an unmanned ship navigation system based on vision and biological monitoring proposed by the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Example 1: Unmanned ship navigation in basic environment

[0049] In relatively simple inland river or lake environments, unmanned ships mainly rely on visual perception modules and basic navigation control modules for navigation.

[0050] Obstacle recognition of the visual perception module:

[0051] Use the camera to obtain the environment image and extract the image features through the multi-scale convolutional neural network (MSCNN). The extracted features are fused using the following formula:

[0052]

[0053] Among them, O represents the output feature after fusion, O l represents the output of the lth layer convolution, α l is the weight coefficient of the lth layer, and L is the number of convolutional layers. This feature fusion technology can weightedly combine image features at different levels to generate more accurate obstacle recognition results.

[0054] Path planning of the navigation control module:

[0055] After obtaining the obstacle position, the navigation control module uses the A* algorithm to calculate the optimal path. The calculation formula is:

[0056] Where L represents the total path length of the route, x i ,y i is the coordinate of the i-th node on the path, m is the number of path nodes, λ is the risk weight coefficient, and R is the risk value detected along the way. The algorithm combines the A* algorithm with the optimization algorithm based on neural networks to calculate the optimal navigation path and dynamically adjust the path to avoid risks. The algorithm ensures that the unmanned ship can bypass the detected obstacles and reach the destination along the shortest path.

[0057] Example 2: Precise navigation in complex waters

[0058] In complex waters such as harbors or estuaries, dynamic changes in the environment increase the difficulty of navigation, so the system requires more advanced modules to work together.

[0059] Collaborative work of the visual perception module and the biological monitoring module:

[0060] The system uses cameras and lidar to detect obstacles, and uses an underwater sonar system to detect underwater creatures. The Bayesian classification algorithm is used to identify the detected creatures, and the classification probability calculation formula is as follows:

[0061]

[0062] Among them, P(c|X) represents the posterior probability of biological category c given feature x, P(X|c) represents the likelihood of feature x under category c, P(c) represents the prior probability of biological category c, and P(X) represents the marginal probability of feature X.

[0063] Risk assessment and path adjustment of the navigation control module:

[0064] The system evaluates the risk along the route and adjusts the route using the following formula:

[0065]

[0066] Among them, R is the comprehensive risk value, r j represents the risk measure of the jth risk source detected, d j is the distance between the unmanned ship and the jth risk source, and k is the number of risk sources. This formula combines distance and risk measurement to ensure that the unmanned ship can effectively avoid targets that are close and have higher risks. The final path planning formula is:

[0067]

[0068] Example 3: Adaptive navigation for emergencies

[0069] During the navigation of the unmanned ship, it may encounter unexpected events, such as the sudden appearance of floating objects or rapidly approaching ships.

[0070] Adaptive adjustment module response:

[0071] When an emergency is detected, the adaptive adjustment module is activated to calculate the adjustment amount of the heading angle. The adjustment formula is:

[0072]

[0073] Among them, Δθ represents the adjustment amount of the heading angle, γ is the adjustment coefficient, is the partial derivative of the risk value with respect to the heading angle. This adjustment mechanism can dynamically adjust the navigation path according to the real-time monitored environment and biological information to ensure that the unmanned ship can reach the target location safely and effectively.

[0074] Real-time path adjustment:

[0075] After adjusting the heading, the navigation control module recalculates the deviation between the current route and the adjusted route, using the dynamic time warping (DTW) algorithm to calculate the deviation:

[0076] DTW(i,j)=min(DTW(i-1,j-1),DTW(i-1,j),DTW(i,j-1))+d(i,j)

[0077] Among them, DTW(i,j) represents the cumulative distance between the i-th point and the j-th point, and d(i,j) represents the local distance between the i-th point and the j-th point. The algorithm can effectively match the actual route with the planned route, evaluate the navigation deviation, and make path corrections when necessary. By adjusting the route, the system ensures that the unmanned ship can avoid sudden dangers and maintain safe navigation.

[0078] Example 4: Application in marine scientific research

[0079] When conducting marine scientific research missions in the open sea or deep sea areas, the system needs to navigate autonomously for a long time and collect marine environmental data.

[0080] Integrated application of visual perception and biological monitoring:

[0081] The system uses low-light sensing cameras and high-precision lidar to identify obstacles in the deep sea, such as floating ice and waves. The sonar system is responsible for detecting the activities of marine life and identifying organisms using Bayesian classification and K-NN algorithms. The algorithm formula is the same as above.

[0082] Dynamic path adjustment and data transfer:

[0083] In bad weather or complex sea conditions, the system adjusts the route in real time and uses the following formula for risk assessment and route optimization:

[0084]

[0085] The communication module transmits real-time data to the remote control center via a satellite link, ensuring that the operating status and environmental data of the unmanned ship can be transmitted back to the scientific research base without delay.

[0086] The above embodiments are specific implementations of the present invention, and those skilled in the relevant technical field can make different changes and modifications based on the content of the present invention without departing from the spirit and scope of the present invention. Therefore, all equivalent implementations that do not depart from the spirit and scope of the present invention should be included in the protection scope of the present invention.

[0087] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in the field. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An unmanned ship navigation system based on vision and biological monitoring, characterized in that: The system comprises: Visual perception module: used to obtain visual information of the surrounding environment of the unmanned ship. The visual perception module includes a high-resolution camera and a laser radar, wherein the camera is used to capture image information of the surrounding environment, and the laser radar is used to measure the distance of the target object and its relative position; Biological monitoring module: used to detect the presence of underwater organisms. The biological monitoring module includes an underwater sonar system and a biological recognition algorithm. The underwater sonar system detects the size, location and type of underwater organisms by emitting sound waves and receiving reflected signals; Navigation control module: used to receive data from the visual perception module and the biological monitoring module, and the navigation control module generates an optimal route through an adaptive navigation algorithm, which performs path planning based on the fused environmental and biological monitoring data; Communication module: used to transmit the location information, environmental data and biological monitoring data of the unmanned ship to the remote control center in real time. The communication module also supports receiving instructions issued by the remote control center.

2. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The visual information obtained by the visual perception module is used to identify obstacles through the following formula: Among them, f(x,y ) represents the feature function of the image, gi(x,y) represents the image features extracted from different scales, and w i is the weight of each feature, and n is the number of features.

3. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The biometric recognition algorithm is based on the Bayesian classifier and is used to calculate the classification probability of the organism. The calculation formula is as follows: Among them, P(c|X) represents the posterior probability of biological category c under the condition of given feature x, P(X|c) represents the likelihood of feature x under category c, P(c) represents the prior probability of biological category c, and P(X) represents the marginal probability of feature X.

4. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The navigation control module calculates the optimal route using the following formula: Where L represents the total path length of the route, x i ,y i is the coordinate of the i-th node on the path, m is the number of path nodes, λ is the risk weight coefficient, and R is the risk value detected along the way.

5. The unmanned ship navigation system based on vision and biological monitoring according to claim 4 is characterized in that: The risk value R is calculated according to the following formula: Among them, r j represents the risk measure of the jth risk source detected, d j is the distance between the unmanned ship and the jth risk source, and k is the number of risk sources.

6. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The visual perception module recognizes obstacles through a multi-scale convolutional neural network, and the output of the network is multi-layered feature fused through the following formula: Among them, O represents the output feature after fusion, O l represents the output of the lth layer convolution, α l is the weight coefficient of the lth layer, and L is the number of convolutional layers.

7. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The biological monitoring module identifies organisms through a classification model based on the K-nearest neighbor algorithm, and its distance measurement formula is: Where D(x,y) represents the Euclidean distance between the biological feature vector x and the sample vector y, and p is the feature dimension. The algorithm calculates the distance between biological features and identifies the most similar biological species, thereby achieving accurate identification of aquatic organisms.

8. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The navigation control module combines the dynamic time warping algorithm to match the actual route of the unmanned ship with the planned route, and the calculation formula is: DTW(i,j)=min(DTW(i-1,j-1),DTW(i-1,j),DTW(i,j-1))+d(i,j) Among them, DTW(i,j) represents the cumulative distance between the i-th point and the j-th point, and d(i,j) represents the local distance between the i-th point and the j-th point. The algorithm can effectively match the actual route with the planned route, evaluate the navigation deviation, and make path corrections when necessary.

9. The unmanned ship navigation system based on vision and biological monitoring according to claim 1 is characterized in that: The system includes an adaptive adjustment module, which adjusts the route of the unmanned ship in real time according to environmental changes and biological behaviors, and the adjustment is based on the following formula: Among them, Δθ represents the adjustment amount of the heading angle, γ is the adjustment coefficient, is the partial derivative of the risk value with respect to the heading angle. This adjustment mechanism can dynamically adjust the navigation path according to the real-time monitored environment and biological information to ensure that the unmanned ship can reach the target location safely and effectively.

10. The unmanned ship navigation system based on vision and biological monitoring according to claim 1, characterized in that: The communication module uses a symmetric encryption algorithm for data transmission, and the encryption process follows the following formula: C=E(K,M)=M⊕H(K) Among them, C is the ciphertext, E(K,M) is the encryption function, M is the plaintext, K is the key, and H(K) is the pseudo-random number generated by the hash function.

Images