Navigation method and system of inland river electronic channel chart based on ship-shore combination

Abstract

The invention discloses a navigation method and system for an inland river electronic channel chart based on ship-shore combination, and belongs to the field of ship navigation. The method comprises the following steps that S1, data of shore-end sensors and ship-end sensors are obtained, the shore-end sensors comprise a radar, a camera, a GPS navigation system and a meteorological and hydrological sensor, and the ship-end sensors comprise a radar, a camera, a GPS navigation system, a sonar and an inertial navigation system; s2, processing the data of the shore-end sensor and the ship-end sensor so as to update an inland river electronic navigation channel chart in real time; s3, navigation is performed based on the inland river electronic navigation channel chart updated in real time. According to the method, the data timeliness and the automation degree of the inland river electronic navigation channel chart are improved, so that various navigation requirements of ships are met, high-reliability support is provided for ship navigation, and the ship navigation safety is ensured.

Description

Technical Field

[0001] The invention belongs to the field of ship navigation, and in particular, relates to a navigation method and system based on a ship-shore combined inland river electronic navigation chart. Background Art

[0002] Electronic navigation charts for inland waterways play a vital role in modern shipping. However, there are some problems in the existing production process of electronic navigation charts. For example, there is a lot of manual participation: the production and update of electronic navigation charts require a lot of manual intervention, which is inefficient; low degree of automation: the lack of automated data collection and processing means leads to untimely updates of navigation chart data; poor data timeliness: the existing system cannot reflect the latest changes in inland waterways in real time, affecting the navigation and safety of ships; insufficient information integration: the hydrological data provided by the shore monitoring system cannot be directly applied to the electronic navigation chart on board, and the information is isolated.

[0003] The comparative document (CN117968648A) discloses a waterway in-situ monitoring system that is adaptive to hydrological conditions, including an intelligent sensing module, a dynamic analysis and decision-making module, an adaptive control module, a real-time communication and feedback module, and an energy management and optimization module, wherein: the intelligent sensing module: is equipped with a variety of hydrological monitoring sensors; the dynamic analysis and decision-making module: uses data analysis technology to analyze the changing trends of hydrological conditions and identify indicators; the adaptive control module: automatically adjusts monitoring strategies or parameters; the real-time communication and feedback module: is responsible for transmitting monitoring data, analysis results and adjusted monitoring strategies to a remote monitoring center; the energy management and optimization module: ensures that the monitoring system operates efficiently under different monitoring strategies.

[0004] However, the above-mentioned comparative documents mainly provide hydrological data monitoring, but fail to achieve real-time integration and annotation of electronic navigation charts for ships' use. Summary of the invention

[0005] The present invention aims to overcome the shortcomings of the prior art and proposes a navigation method and system based on a ship-shore combined inland waterway electronic navigation chart to achieve the following purposes: to improve the data timeliness and automation of the inland waterway electronic navigation chart, thereby meeting the diverse navigation needs of ships, providing high-reliability support for ship navigation, and thus ensuring the safety of ship navigation.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: a navigation method based on an inland river electronic navigation chart combined with a ship and a shore, the method comprising the following steps:

[0007] S1. Acquire data from shore-side and ship-side sensors, wherein the shore-side sensors include radar, camera, GPS navigation system, meteorological and hydrological sensors, and the ship-side sensors include radar, camera, GPS navigation system, sonar, and inertial navigation system;

[0008] S2, processing the data of the shore-side and ship-side sensors to update the inland river electronic navigation chart in real time;

[0009] S3. Navigation based on real-time updated inland waterway electronic navigation charts.

[0010] Preferably, in step S2, the YOLOv10 algorithm is used to perform target detection on the images captured by the ship-side and shore-side cameras, and the objects of target detection include ships, channel signs, buildings, floating objects on the water, and weather phenomena.

[0011] Preferably, in step S2, for similar data available from both shore-side and ship-side sensors, a Kalman filter is used to perform data fusion to achieve ship-shore integration, and the similar data include radar data from the ship-side and shore-side, and GPS data from the ship-side and shore-side.

[0012] Preferably, in step S3, navigating based on the real-time updated inland river electronic navigation chart includes performing water depth measurement based on sonar data, and an adaptive beamforming technology is used in the water depth measurement process, that is, dynamically adjusting the weight vector w of the receiving array to maximize the signal-to-noise ratio of the signal, that is:

[0013] w=R -1 a(θ),R=E[x(t)x H (t)],

[0014] Where R is the covariance matrix of the received signal; x(t) represents the signal vector of the receiving array; t represents time; the superscript H represents conjugation, and a(θ) is the steering vector in the desired direction θ, which represents the phase difference of each receiving array when the signal arrives from this direction;

[0015] The optimization goal of the weight vector w is to satisfy w H a(θ)=1minimizes the output signal power, i.e., min w w H Rw, by solving the optimization objective, the optimal weight vector is obtained and used for signal beamforming processing.

[0016] Preferably, in step S3, water depth measurement is performed based on sonar data, including depth calculation by least square method combined with time reversal technology, that is, assuming that the time from the sonar emission to the reception of the echo is t, the propagation speed is v, and the propagation distance is d, The echo data (t i ,d i ) is fitted, and the residual sum of squares S is minimized to solve the best fitting linear equation to obtain the optimal water depth estimation value d, that is:

[0017]

[0018] In the solution process, time reversal processing is introduced, and the formula is:

[0019]

[0020] Where h(-τ) is the time-reversed impulse response of the system, x(t) is the received signal, and y(t) is the output signal after time reversal.

[0021] Preferably, in step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes waterway collision warning. Specifically, the waterway collision warning includes the following steps:

[0022] S311. Obtain the data of the radar and camera after the ship-shore combination, use the weighted fusion method to unify the radar and camera data in a global coordinate system, and define the weighted fusion formula of the radar and camera as follows:

[0023] P fused =αP rader +(1+α)P camera ,

[0024] Among them, P fused represents the fused target position vector, P rader is the target position vector detected by the radar, P camera is the target position vector recognized by the camera, α is the weight coefficient, which depends on the reliability of radar and camera data;

[0025] S312, calculate the risk of ship collision based on the closest point in time TCPA) and the closest point in distance DCPA, when DCPA i (t) is less than the set safety distance threshold D th , and TCPA i (t) If it is within the preset time range, it is considered that there is a collision risk and a warning is triggered. The calculation formula is as follows:

[0026]

[0027] Among them, D i (t) indicates the distance of other vessels or objects relative to own ship; v rel,i (t) represents the relative speed of other ships or objects relative to the own ship; i is used to identify other ships or objects; P(t) represents the position of the own ship, P i (t) Indicates the position of other vessels or objects.

[0028] Preferably, in step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes a waterway deviation warning. Specifically, the waterway deviation warning includes the following steps:

[0029] S321, obtaining GPS data and inertial navigation system data after ship-shore integration to calculate the current position of the ship;

[0030] S322: Calculate the lateral offset Δd(t) of the ship relative to the center of the channel according to the current position of the ship. If the offset Δd(t) exceeds the preset safety threshold d th , then the ship is judged to have the risk of deviating from the course and triggers the deviation warning. The deviation calculation formula is as follows:

[0031]

[0032] Where C(t) = [c x (t),c y (t)] represents the coordinate of the channel centerline closest to the ship; P(t) = |p x (t),p y (t)| represents the current position coordinates of the ship, the subscript x represents the horizontal direction, and the subscript y represents the vertical direction.

[0033] Preferably, in step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes flow and congestion monitoring. Specifically, the flow and congestion monitoring includes the following steps:

[0034] S331, obtain the real-time position of each ship on the route through GPS and inertial navigation system data At the same time, the real-time speed of each ship is obtained through the combined radar data of the ship and the shore, and the channel is divided into several gridded areas A k , each region covers a geographic space of M×N;

[0035] S332. Use the find function to map the position of the ship to the corresponding grid, i.e. A k =find(P i (t)); then within time t, area A k Channel flow Q k (t) is expressed as Among them, N k (t) is the time passing through area A k The number of ships;

[0036] S333. For ships in the same grid, calculate the distance between them When ij(t) is less than the preset safety distance d safe When , it is judged that there is a congestion risk in the area; define the congestion index C k (t) is area A k The distance between ships is less than d safe The ratio of the number of ship pairs to the total number of ship pairs, that is:

[0037]

[0038] When C k (t) Exceeding the safety distance C th When , it is determined that there is congestion in the area.

[0039] Preferably, in step S3, navigation based on the real-time updated inland river electronic navigation chart includes ship path planning. Specifically, the ship path planning includes the following steps:

[0040] S341, obtaining data from different sensors and performing preprocessing;

[0041] S342. Use the deep Q learning algorithm to build a path planning model for path planning. The input of the algorithm is the preprocessed real-time sensor data and historical sensor data, which is expressed by the following formula:

[0042] Q(s t ,a t )=Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a t+1 )-Q(s t ,α t )),

[0043] Among them, Q(s t ,a t ) indicates state s t Take action a t Q value, r t is the reward, α is the learning rate, γ is the discount factor, and t is the time;

[0044] S343, for each input data of the path planning model, by adjusting the weight coefficient w of the input data i In order to optimize the model, the adjustment of the weight parameters refers to the change trend of the corresponding input data, wherein the change trend is reflected by constructing a time series analysis ARIMA model, and the ARIMA model is expressed by the following formula:

[0045] Y t =c+φ1Y t-1+φ2Y t-2 +…+θ1ε t-1 +θ2ε t-2 +…+ε t ,

[0046] Among them, Y t is the current value of the corresponding input data, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the noise, and t represents the current time.

[0047] S344, based on the algorithm analysis, output the corresponding decision results, including the recommended optimal waterway safety suggestions; at the same time, establish a feedback mechanism to collect feedback from operators and actual application effects, and adjust the weights in real time based on the feedback, i.e., w t+1 =w t +β·feedback(t), where β is the learning rate of feedback, feedback(t) is the feedback value collected within time t, and w t is the current weight, w t+1 is the updated weight.

[0048] At the same time, the present application also proposes a navigation system based on a ship-shore integrated inland waterway electronic navigation chart, the system is constructed according to the above-mentioned navigation method based on a ship-shore integrated inland waterway electronic navigation chart, the system includes a ship-side sensor, a shore-side sensor, a ship-side controller, and a remote communication unit, wherein the ship-side sensor is connected to the ship-side controller, the shore-side sensor is connected to the ship-side controller through the remote communication unit, and the ship-side controller is used to navigate according to the above-mentioned navigation method based on a ship-shore integrated inland waterway electronic navigation chart.

[0049] The technical effects of the present invention are:

[0050] (1) Real-time data collection and processing are achieved through the combination of ship-side and shore-side, ensuring that the data of the electronic navigation chart is updated in a timely manner.

[0051] (2) The entire method process realizes the automatic update of the inland waterway electronic navigation chart by combining a controller with multiple sensors, which greatly reduces the manual participation and improves the work efficiency and the degree of automation of data processing.

[0052] (3) This application improves the accuracy and reliability of data through multi-sensor data fusion and advanced algorithm processing, thereby ensuring the reliability of inland waterway electronic navigation charts.

[0053] (4) Inland waterway electronic navigation charts can adapt to the complex and changing environment of inland waterways and provide more accurate and real-time channel information, thereby providing reliable decision-making support for ship navigation and meeting the diverse needs of ship navigation, such as ship collision avoidance and optimization of navigation paths, thereby improving navigation safety and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A flow chart of a navigation method based on a ship-shore combined inland river electronic navigation chart according to an embodiment of the present invention;

[0055] Figure 2 The present invention is a structural diagram of a navigation system based on a ship-shore combined inland waterway electronic navigation chart according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following is a further detailed description of the specific implementation of the present invention through the description of the embodiments with reference to the accompanying drawings, with the purpose of helping 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 invention and to facilitate its implementation. In order to make the technical solution of the present invention clearer, the present invention is explained through the following embodiments:

[0057] A navigation method based on ship-shore combined inland waterway electronic navigation chart, such as Figure 1 As shown, the method comprises the following steps:

[0058] S1. Acquire data from shore-side and ship-side sensors; the shore-side sensors include radar, camera, GPS navigation system, meteorological and hydrological sensors, and the ship-side sensors include radar, camera, GPS navigation system, sonar, and inertial navigation system;

[0059] S2, processing the data of the shore-side and ship-side sensors to update the inland river electronic navigation chart in real time;

[0060] S3. Navigation based on real-time updated inland waterway electronic navigation charts.

[0061] Specifically, in step S2, the YOLOv10 algorithm is used to detect targets in the images captured by the ship-side and shore-side cameras. The objects of target detection include ships, channel signs, buildings, floating objects on the water, weather phenomena, etc. The specific detection targets are classified as follows:

[0062] Other vessels: Identify other nearby vessels of various types including cargo ships, passenger ships, tugboats, small craft, etc. and determine their size, position, speed and direction;

[0063] Channel markings: Identify fixed or floating channel markings such as buoys, signs, and beacons and their relative positions, and interpret the marking information (such as channel direction, shallow water warning, etc.);

[0064] Buildings: Identify bridges, docks, piers and other artificial or natural obstacles across the waterway, as well as their distance, safe height for passing, width and other information;

[0065] Floating objects on the water: Identify the location, size and shape of floating objects on the water, such as trees, garbage, driftwood, etc.;

[0066] Severe weather phenomena: Identify small risk areas with poor visual visibility due to heavy rain, haze, ice and snow, and recommend slowing down or changing routes based on meteorological sensor data.

[0067] The YOLOv10 algorithm is an existing target detection algorithm with the advantages of high detection accuracy and efficiency. In this embodiment, the image data captured by the camera is processed by the YOLOv10 algorithm to ensure the reliability and accuracy of the data. The specific process is as follows:

[0068] (1) Input image: Input the image sequence captured by the camera. The image size is W*H*C, where W and H are the width and height of the image respectively, and C is the number of channels, which is RGB channels in this case.

[0069] (2) Feature extraction: The feature map of the image is extracted through a convolutional neural network (CNN). This step includes multiple convolutional layers and pooling layers.

[0070] (3) Bounding box prediction and classification: For each pixel in the feature map, predict the bounding box of the target (x, y, w, h), where x and y are the center coordinates, and w and h are the width and height. At the same time, the network outputs the target category c within each bounding box. The target category is as mentioned above.

[0071] (4) Non-maximum suppression (NMS): Non-maximum suppression is used to eliminate redundant bounding boxes, ensuring that only one detection box is retained for each object. The algorithm uses the following formula to calculate the confidence score s of each bounding box: s = p(c) × IoU(b,b true ), where p(c) is the class probability, IoU(b,b true ) is the intersection-over-union ratio of the predicted box and the true box, b represents the predicted box, b true Represents the true box.

[0072] (5) Target tracking: For continuous video frames, SORT (Simple Online and Realtime Tracking) is used to track the detected targets. The core idea of ​​the SORT algorithm is to use the Kalman filter for state estimation and solve the data association problem through the Hungarian algorithm to achieve real-time tracking of multiple targets. In the SORT algorithm, the Kalman filter is first initialized, and then the predicted state is corrected through new observation data. SORT uses the Hungarian algorithm for data association, that is, to solve the matching problem between new detection results and existing tracked targets. The calculation of data association includes the calculation of the cost matrix and the solution of the Hungarian algorithm. The cost calculation calculates the cost between each new detected target i and the existing tracked target j, C(i,j)=1-IoU(B i ,B j ), where B i and B j They are the bounding boxes of the detected target and the tracked target, respectively. The Hungarian algorithm can minimize the cost matrix and find the optimal match. The matching result will determine the association between the new detected target and the existing tracked target. For successfully matched targets, the Kalman filter update step is used to update its state. If a detection result does not match any tracked target, a new Kalman filter is initialized to track the target. If a tracked target fails to match a new detection result for several consecutive frames, it is considered that the target has been lost and tracking is stopped. Finally, SORT outputs the status of all tracked targets, including the bounding box position and identifier of each target, which can be used by subsequent algorithms.

[0073] In step S2 of this embodiment, for similar data of both shore-side and ship-side sensors, Kalman filter is used for data fusion to realize ship-shore integration, and the similar data includes radar data of the ship-side and shore-side, GPS data of the ship-side and shore-side, etc. Specifically, the shore-side radar data can provide the overall situation of the river, such as the width of the channel, the shape of the river bank, and the detection of obstacles along the way. The ship-side radar data is mainly used for close-range detection in front of the ship, detecting other ships, buoys, bridges and other objects. This embodiment fuses the shore-side radar data and the ship-side radar data, which can improve the accuracy of obstacle detection and the accuracy of channel monitoring. At the same time, it can supplement more extensive environmental information within the local field of view of the ship to form a more comprehensive environmental perception. The shore-side GPS data is used for the positioning and reference of the shore-side equipment, and the ship-side GPS data provides the precise position of the ship. This embodiment can further improve the positioning accuracy by comparing and fusing the ship-side GPS data with the shore-side GPS data, ensuring that the ship's position matches the reference point on the shore.

[0074] In the fusion of the same type of data, the application of Kalman filter can effectively integrate measurement data from different sources, reduce noise and optimize data quality. The fusion steps and formulas are as follows:

[0075] State estimation: Set the data state at each moment to x k ;

[0076] Prediction step: x k∣k-1 =Ax k-1∣k-1 +Bu k ,P k∣k-1 =AP k-1∣k-1 A T +Q; where A is the state transfer matrix, P k∣k-1 is the prediction covariance matrix, and Q is the process noise covariance matrix.

[0077] Update steps:

[0078] K k =P k∣k-1 H T (HP k∣k-1 H T +R) -1 ,

[0079] x k∣k =x k∣k-1 +K k (z k -Hx k∣k-1 ),

[0080] P k∣k =(IK k H)P k∣k-1 ;

[0081] Among them, K k is the Kalman gain matrix, H is the observation matrix, R is the observation noise covariance matrix, z k is the fused observation data.

[0082] During the data fusion process, the Kalman filter will perform a weighted average of similar data from the ship and shore ends, and use the estimated Kalman gain to correct the predicted value, making the final fusion result closer to the actual situation. Through multiple iterations, the noise and error are gradually reduced, and the real-time performance and accuracy of the system are improved.

[0083] Further, in step S3 of this embodiment, navigation based on the real-time updated inland river electronic navigation chart includes performing water depth measurement based on sonar data, and adaptive beamforming technology is used in the water depth measurement process, that is, dynamically adjusting the weight vector w of the receiving array to maximize the signal-to-noise ratio of the signal, that is:

[0084] w=R-1 a(θ),R=E[x(t)x H (t)],

[0085] Where R is the covariance matrix of the received signal; x(t) represents the signal vector of the receiving array; t represents time; the superscript H represents conjugation, and a(θ) is the steering vector in the desired direction θ, which represents the phase difference of each receiving array when the signal arrives from this direction. Compared with traditional delay-and-sum beamforming, adaptive beamforming can more effectively suppress interference and noise and provide higher spatial resolution and accuracy.

[0086] The optimization goal of the weight vector w is to satisfy w H a(θ)=1minimizes the output signal power, i.e., min w w H Rw, by solving the optimization objective, the optimal weight vector is obtained and used for signal beamforming processing.

[0087] Then, in the water depth measurement based on sonar data, the depth calculation is also carried out by combining the least square method with the time reversal technology, that is, assuming that the time from the sonar emission to the reception of the echo is t, the propagation speed is v, and the propagation distance is d, The echo data (t i ,d i ) is fitted, and the residual sum of squares S is minimized to solve the best fitting linear equation to obtain the optimal water depth estimation value d, that is:

[0088]

[0089] Time reversal technology can improve measurement accuracy by inverting the received signal waveform and amplifying the signal response at a specific time point. In a multipath propagation environment, time reversal technology can enhance the direct path signal and suppress the interference caused by reflection and scattering, thereby accurately estimating the depth of the target. Therefore, this embodiment introduces time reversal processing in the solution process, and its formula is:

[0090]

[0091] Among them, h(-τ) is the time-reversed impulse response of the system, x(t) is the received signal, and y(t) is the output signal after time reversal. After time reversal processing, the focusing effect of the obtained echo signal at a specific depth will be more obvious, thereby improving the accuracy of water depth measurement. These methods can provide stable and reliable water depth data in complex inland river environments, providing important support for ship navigation and channel monitoring.

[0092] In step S3 of this embodiment, navigation based on the real-time updated inland waterway electronic navigation chart includes waterway collision warning. Specifically, the waterway collision warning includes the following steps:

[0093] S311. Obtain the data of the radar and camera after the ship-shore combination, use the weighted fusion method to unify the radar and camera data in a global coordinate system, and define the weighted fusion formula of the radar and camera as follows:

[0094] P fused =αP rader +(1+α)P camera ,

[0095] Among them, P fused represents the fused target position vector, P rader is the target position vector detected by the radar, P camera is the target position vector recognized by the camera, α is the weight coefficient, which depends on the reliability of radar and camera data;

[0096] S312, calculate the risk of ship collision based on the closest point in time TCPA) and the closest point in distance DCPA, when DCPA i (t) is less than the set safety distance threshold D th , and TCPA i (t) Within the preset time range (the time set in this embodiment is 5 minutes, which can be flexibly set according to actual conditions), it is considered that there is a collision risk and an early warning is triggered. The calculation formula is as follows:

[0097]

[0098] Among them, D i (t) indicates the distance of other vessels or objects relative to own ship; v rel,i (t) represents the relative speed of other ships or objects relative to the own ship; i is used to identify other ships or objects; P(t) represents the position of the own ship, P i (t) Indicates the position of other vessels or objects.

[0099] In step S3, navigation based on the real-time updated inland waterway electronic navigation chart also includes a waterway deviation warning. Specifically, the waterway deviation warning includes the following steps:

[0100] S321, obtaining GPS data and inertial navigation system data after ship-shore integration to calculate the current position of the ship;

[0101] S322, based on the current position of the ship, calculate the lateral offset Δd(t) of the ship relative to the center of the channel. The center line of the channel is pre-calibrated. If the offset Δd(t) exceeds the preset safety threshold d th , then the ship is judged to have the risk of deviating from the course and triggers the deviation warning. The deviation calculation formula is as follows:

[0102]

[0103] Where C(t) = [c x (t),c y (t)] represents the coordinate of the channel centerline closest to the ship; P(t) = |p x (t),p y (t)| represents the current position coordinates of the ship, the subscript x represents the horizontal direction, and the subscript y represents the vertical direction.

[0104] In step S3, navigation based on the real-time updated inland waterway electronic navigation chart also includes flow and congestion monitoring. Specifically, the flow and congestion monitoring includes the following steps:

[0105] S331, obtain the real-time position of each ship on the route through GPS and inertial navigation system data At the same time, the real-time speed of each ship is obtained through the combined radar data of the ship and the shore, and the channel is divided into several gridded areas A k , each region covers a geographic space of M×N;

[0106] S332. Use the find function to map the position of the ship to the corresponding grid, i.e. A k =find(P i (t)); then within time t, area A k Channel flow Q k (t) is expressed as Among them, N k (t) is the time passing through area A k The number of ships;

[0107] S333. For ships in the same grid, calculate the distance between them When ij (t) is less than the preset safety distance d safe When , it is judged that there is a congestion risk in the area; define the congestion index C k (t) is area A k The distance between ships is less than d safe The ratio of the number of ship pairs to the total number of ship pairs, that is:

[0108]

[0109] When C k (t) Exceeding the safety distance C th When the ship is in a traffic jam, it is judged that there is congestion in the area. Then, a warning mark can be formed on the electronic navigation chart and information can be generated to suggest that the ship change the speed or route.

[0110] In step S3, navigation based on the real-time updated inland river electronic navigation chart also includes ship path planning, which is analyzed by integrating real-time data and historical data to ensure the safety and efficiency of navigation. Specifically, the ship path planning includes the following steps:

[0111] S341, acquiring data from different sensors and performing preprocessing, wherein the preprocessing includes operations such as data cleaning and normalization;

[0112] S342. Use a deep Q learning algorithm (the deep Q learning algorithm is a prior art and will not be described in detail in the present invention) to build a path planning model for path planning. The input of the algorithm is the preprocessed real-time sensor data and historical sensor data, which is expressed by the following formula:

[0113] Q(s t ,a t )=Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a t+1 )-Q(s t ,a t )),

[0114] Among them, Q(s t ,a t ) indicates state s t Take action a t Q value, r t is the reward, α is the learning rate, γ is the discount factor, and t is the time;

[0115] S343, for each input data of the path planning model, by adjusting the weight coefficient w of the input data iTo optimize the model. The weight adjustment mechanism can dynamically adjust the weight of the parameters according to environmental conditions, historical data analysis, etc. Environmental conditions such as: when sailing at night, the weight of radar data can be increased; the weight of the impact of sunny days on weather can be reduced. The adjustment of weight parameters refers to the change trend of the corresponding input data. By analyzing historical data, the system can identify the long-term impact trend of each parameter and automatically optimize the weight. Among them, the change trend is reflected by constructing a time series analysis ARIMA model, and the ARIMA model is expressed by the following formula:

[0116] Y t =c+φ1Y t-1 +φ2Y t-2 +…+θ1ε t-1 +θ2ε t-2 +…+ε t ,

[0117] Among them, Y t is the current value of the corresponding input data, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the noise, and t represents the current time.

[0118] S344. According to the algorithm analysis, the corresponding decision results are output, including the recommended optimal waterway safety suggestions; at the same time, in order to improve the accuracy and practicality of the algorithm, a feedback mechanism is established to collect feedback from operators and actual application effects, and the weights are adjusted in real time according to the feedback, that is, w t+1 =w t +β·feedback(t), where β is the learning rate of feedback, feedback(t) is the feedback value collected within time t, and w t is the current weight, w t+1 This embodiment ensures the adaptability of the algorithm model to the actual scenario by regularly retraining and optimizing the algorithm model based on the collected feedback data.

[0119] In summary, the present invention, based on the real-time updated inland waterway electronic navigation chart, meets various navigation needs, such as channel deviation and collision warning, flow and congestion monitoring, ship path planning, etc., improves the reliability of navigation, and ensures the navigation safety of ships.

[0120] The present invention can provide instant feedback and customized monitoring through ship-side equipment, and can provide a global perspective and long-term monitoring through shore-side equipment. The combination of the two can achieve more accurate channel monitoring and navigation support. At the same time, the ship-side equipment adapts to complex environments, and the shore-side equipment covers a comprehensive area. Together, they can cope with the challenges and changes of various inland waterways. Based on the updated inland electronic waterway chart after integrating shore-side and ship-side data, the present invention can provide comprehensive waterway status analysis and decision support, which helps to optimize navigation paths and reduce accident risks.

[0121] At the same time, the present application also proposes a navigation system based on a ship-shore combined inland river electronic navigation chart, the system is constructed according to the above-mentioned navigation method based on a ship-shore combined inland river electronic navigation chart, such as Figure 2 As shown, the system includes a ship-side sensor, a shore-side sensor, a ship-side controller, and a remote communication unit, wherein the ship-side sensor is connected to the ship-side controller, the shore-side sensor is connected to the ship-side controller through the remote communication unit, and the ship-side controller is used to navigate according to the above-mentioned navigation method based on the ship-shore combined inland waterway electronic navigation chart. The remote communication unit of this embodiment can be implemented by radio technology or the Internet and other technologies. The ship-side controller can use control devices such as CPU and PLC, which can be flexibly selected according to actual conditions.

[0122] The shore sensors include radar, camera, GPS navigation system, meteorological and hydrological sensors, etc., among which the shore radar can be installed on the river bank or high buildings, the shore camera can be installed on bridges, shores or floating platforms, and other sensors such as meteorological and hydrological sensors can be installed at locations close to the waterway but do not affect navigation. All shore sensors are connected to the ship controller through the remote communication unit. In other feasible embodiments of the present invention, the shore sensor data can also be processed by the shore data processing unit or controller and then sent to the ship controller to reduce the data processing pressure of the ship controller, thereby improving the data processing speed.

[0123] The ship-side sensors include radar, camera, GPS navigation system, sonar, and inertial navigation system, wherein the ship-side radar and camera can be installed on the mast or hull of the ship, the GPS navigation system can be installed on the mast or superstructure of the ship, the sonar can be installed on the bottom of the hull or the stern; the inertial navigation system can be installed at the center or center of gravity of the ship. All ship-side sensors can be connected to the ship-side controller through data buses such as Ethernet bus and CAN bus.

[0124] The present invention is described above by way of example in conjunction with the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-mentioned methods. As long as various non-substantial improvements are made using the method concept and technical solution of the present invention; or the above concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A navigation method based on a ship-shore combined inland river electronic navigation chart, characterized in that: The method comprises the following steps: S1. Acquire data from shore-side and ship-side sensors, wherein the shore-side sensors include radar, camera, GPS navigation system, meteorological and hydrological sensors, and the ship-side sensors include radar, camera, GPS navigation system, sonar, and inertial navigation system; S2, processing the data of the shore-side and ship-side sensors to update the inland river electronic navigation chart in real time; S3. Navigation based on real-time updated inland waterway electronic navigation charts.

2. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 1 is characterized by: In step S2, the YOLOv10 algorithm is used to detect targets in the images captured by the ship-side and shore-side cameras. The objects of target detection include ships, channel signs, buildings, floating objects on the water, and weather phenomena.

3. A navigation method based on ship-shore combined inland river electronic navigation chart according to claim 1 or 2, characterized in that: In step S2, for similar data available from both the shore and ship sensors, a Kalman filter is used to perform data fusion to achieve ship-shore integration, and the similar data include radar data from the ship and shore, GPS data from the ship and shore, and image feature data collected by cameras from the ship and shore.

4. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 1 is characterized by: In step S3, navigation based on the real-time updated inland river electronic navigation chart includes water depth measurement based on sonar data, and adaptive beamforming technology is used in the water depth measurement process, that is, the weight vector w of the receiving array is dynamically adjusted to maximize the signal-to-noise ratio of the signal, that is: w=R -1 a(θ),R=E[x(t)x H (t)], Where R is the covariance matrix of the received signal; x(t) represents the signal vector of the receiving array; t represents time; the superscript H represents conjugation, and a(θ) is the steering vector in the desired direction θ, which represents the phase difference of each receiving array when the signal arrives from this direction; The optimization goal of the weight vector w is to satisfy w H a(θ)=1minimizes the output signal power, i.e., min w w H Rw, by solving the optimization objective, the optimal weight vector is obtained and used for signal beamforming processing.

5. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 4 is characterized by: In step S3, water depth measurement is performed based on sonar data, and depth calculation is also performed by least square method combined with time reversal technology, that is: Assume that the time from the sonar emission to the reception of the echo is t, the propagation speed is v, and the propagation distance is The echo data (t i ,d i ) is fitted, and the residual sum of squares S is minimized to solve the best fitting linear equation to obtain the optimal water depth estimation value d, that is: In the solution process, time reversal processing is introduced, and the formula is: Where h(-τ) is the time-reversed impulse response of the system, x(t) is the received signal, and y(t) is the output signal after time reversal.

6. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 3 is characterized by: In the step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes waterway collision warning. Specifically, the waterway collision warning includes the following steps: S311. Obtain the data of the radar and camera after the ship-shore combination, use the weighted fusion method to unify the radar and camera data in a global coordinate system, and define the weighted fusion formula of the radar and camera as follows: P fused =αP rader +(1+α)P camera , Among them, P fused represents the fused target position vector, P rader is the target position vector detected by the radar, P camera is the target position vector recognized by the camera, α is the weight coefficient, which depends on the reliability of radar and camera data; S312, calculate the risk of ship collision based on the closest point in time TCPA) and the closest point in distance DCPA, when DCPA i (t) is less than the set safety distance threshold D th , and TCPA i (t) If it is within the preset time range, it is considered that there is a collision risk and a warning is triggered. The calculation formula is as follows: Among them, D i (t) indicates the distance of other vessels or objects relative to own ship; v rel,i (t) represents the relative speed of other ships or objects relative to the own ship; i is used to identify other ships or objects; P(t) represents the position of the own ship, P i (t) Indicates the position of other vessels or objects.

7. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 3 is characterized by: In the step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes a waterway deviation warning. Specifically, the waterway deviation warning includes the following steps: S321, obtaining GPS data and inertial navigation system data after ship-shore integration to calculate the current position of the ship; S322: Calculate the lateral offset Δd(t) of the ship relative to the center of the channel according to the current position of the ship. If the offset Δd(t) exceeds the preset safety threshold d th , then the ship is judged to have the risk of deviating from the course and triggers the deviation warning. The deviation calculation formula is as follows: Where C(t) = [c x (t),c y (t)] represents the coordinate of the channel centerline closest to the ship; P(t) = |p x (t),p y (t)| represents the current position coordinates of the ship, the subscript x represents the horizontal direction, and the subscript y represents the vertical direction.

8. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 3 is characterized by: In the step S3, navigation based on the real-time updated inland waterway electronic navigation chart includes flow and congestion monitoring. Specifically, the flow and congestion monitoring includes the following steps: S331, obtain the real-time position of each ship on the route through GPS and inertial navigation system data At the same time, the real-time speed of each ship is obtained through the combined radar data of the ship and the shore, and the channel is divided into several gridded areas A k , each region covers a geographic space of M×N; S332. Use the find function to map the position of the ship to the corresponding grid, i.e. A k =find(P i (t)); then within time t, area A k Channel flow Q k (t) is expressed as Among them, N k (t) is the time passing through area A k The number of ships; S333. For ships in the same grid, calculate the distance between them When ij (t) is less than the preset safety distance d safe When , it is judged that there is a congestion risk in the area; define the congestion index C k (t) is area A k The distance between ships is less than d safe The ratio of the number of ship pairs to the total number of ship pairs, that is: When C k (t) Exceeding the safety distance C th When , it is determined that there is congestion in the area.

9. The navigation method based on the ship-shore combined inland river electronic navigation chart according to claim 3 is characterized by: In step S3, navigation based on the real-time updated inland river electronic navigation chart includes ship path planning. Specifically, the ship path planning includes the following steps: S341, obtaining data from different sensors and performing preprocessing; S342. Use the deep Q learning algorithm to build a path planning model for path planning. The input of the algorithm is the preprocessed real-time sensor data and historical sensor data, which is expressed by the following formula: Q(s t ,a t )=Q(s t ,a t )+α(r t +γmax a Q(s t+1 ,a t+1 )-Q(s t ,a t )), Among them, Q(s t ,a t ) indicates state s t Take action a t Q value, r t is the reward, α is the learning rate, γ is the discount factor, and t is the time; S343, for each input data of the path planning model, by adjusting the weight coefficient w of the input data i In order to optimize the model, the adjustment of the weight parameters refers to the change trend of the corresponding input data, wherein the change trend is reflected by constructing a time series analysis ARIMA model, and the ARIMA model is expressed by the following formula: Y t =c+φ1Y t-1 +φ2Y t-2 +…+θ1ε t-1 +θ2ε t-2 +…+e t , Among them, Y t is the current value of the corresponding input data, φ is the autoregressive coefficient, θ is the moving average coefficient, and ε t is the noise, and t represents the current time. S344, according to the algorithm analysis, output the corresponding decision results, including the recommended optimal waterway safety suggestions; at the same time, establish a feedback mechanism to collect feedback from operators and actual application effects, and adjust the weights in real time according to the feedback, that is, w t+1 =w t +β·feedback(t), where β is the learning rate of feedback, feedback(t) is the feedback value collected within time t, and w t is the current weight, w t+1 is the updated weight.

10. A navigation system based on ship-shore combined inland river electronic navigation chart, characterized in that: The system is constructed according to the navigation method based on the ship-shore integrated inland waterway electronic navigation chart as described in any one of claims 1-9, and the system includes a ship-side sensor, a shore-side sensor, a ship-side controller, and a remote communication unit, wherein the ship-side sensor is connected to the ship-side controller, and the shore-side sensor is connected to the ship-side controller via the remote communication unit, and the ship-side controller is used for navigation according to the navigation method based on the ship-shore integrated inland waterway electronic navigation chart as described in any one of claims 1-9.

Images