Electronic chart navigation element rendering system and platform based on graphics acceleration

Abstract

The application relates to the technical field of data processing, and particularly discloses an electronic channel chart navigation element rendering system and platform based on graphics acceleration, which comprises an acquisition module, a sequence module and a write-back module, and a rendering module. The acquisition module divides a target channel section into multiple target river channels, acquires target river channel data of each target river channel, and determines river channel hydrology prediction models of multiple river channel types. The sequence module acquires real-time water level monitoring data of each target river channel, and determines a write-back time sequence of each target river channel. The write-back module acquires electronic channel data, and determines updated channel data of each adjustment time point or interpolation time point in the write-back time sequence. The rendering module performs graphics acceleration-based rendering on the updated channel data of each adjustment time point or interpolation time point in the write-back time sequence, and determines corresponding updated channel charts. The application can realize time sequence dynamic water depth prediction, standardized data write-back, and link closed loop of full-element graphics acceleration rendering, and improves the real-time performance, accuracy and running fluency of the electronic channel chart.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a rendering system and platform for navigation elements of electronic navigation charts based on graphics acceleration. Background Technology

[0002] Electronic navigation chart technology originated before 2000, initially relying mainly on CCD scanning and rasterization of paper-based navigation charts, which could only achieve static visualization and lacked dynamic interaction and navigation capabilities. From 2000 to 2015, it entered a period of rapid development towards vectorization, with the domestic CJ-57 industry standard being established simultaneously, enabling topological storage and attribute querying of navigation elements, and beginning to integrate real-time water level data for simple water depth correction. After 2015, it evolved towards intelligent multi-source fusion, but existing technology still has core shortcomings: water depth updates rely on manual intervention and are highly delayed; dynamically fused data is displayed as an independent layer for secondary overlay; traditional rendering of high-density elements has low frame rates and causes lag on mobile devices; and hydrological prediction is disconnected from navigation chart rendering, failing to meet the real-time safety requirements for inland waterway navigation.

[0003] Therefore, this invention proposes an electronic navigation chart navigation element rendering system and platform based on graphics acceleration. Summary of the Invention

[0004] This invention provides a graphics-accelerated electronic navigation chart navigation element rendering system and platform. The system divides a target navigation section into multiple target rivers at set intervals, acquires and analyzes target river data for each river, determines river hydrological prediction models for multiple river types, acquires real-time water level monitoring data for each target river, determines the dynamic water depth field and status label for each prediction time point in the initial time series of the target river, determines the write-back time series for each target river, acquires electronic navigation chart data based on a standard data format, determines the updated navigation data for each adjustment or interpolation time point in the write-back time series, performs graphics-accelerated rendering on the updated navigation data for each adjustment or interpolation time point in the write-back time series, and determines the updated navigation chart for each adjustment or interpolation time point in the write-back time series. It can solve the problems of incompatibility between dynamic data and standard formats, lag in static updates of electronic navigation charts, and low rendering efficiency. It realizes a closed-loop system with a full link, from segmented hydrological modeling of waterways, time-series dynamic water depth prediction, standardized data write-back to accelerated rendering of all elements of graphics, improving the real-time performance, accuracy and smooth operation of electronic navigation charts, and providing precise support for the entire process of inland waterway navigation safety.

[0005] This invention provides a graphics-accelerated electronic navigation chart navigation element rendering system, comprising: Acquisition module: Divides the target waterway section into multiple target channels according to a set interval, acquires and analyzes the target channel data of each target channel, and determines the channel hydrological prediction model for multiple channel types; Sequence Module: Acquires real-time water level monitoring data for each target river channel. Based on the target river channel data, real-time water level monitoring data, and river hydrological prediction models for all river types, it determines the dynamic water depth field and status label for each prediction time point in the initial time series of the target river channel, and determines the write-back time series for each target river channel. Write-back module: Acquires electronic channel data based on standard data format from electronic channel charts, writes back the dynamic depth field of each adjustment time point or interpolation time point in the write-back time series based on standard data format, and determines the updated channel data for each adjustment time point or interpolation time point in the write-back time series; Rendering module: Performs graphics-accelerated rendering of the updated channel data for each adjustment or interpolation time point in the write-back time series, and determines the updated channel map for each adjustment or interpolation time point in the write-back time series.

[0006] Preferably, the electronic navigation chart navigation element rendering system based on graphics acceleration divides the target navigation section into multiple target waterways according to a set interval, and obtains the target waterway data for each target waterway, wherein the target waterway data includes waterway type, waterway center coordinates, historical hydrological monitoring data and riverbed evolution data. Feature extraction unit: extracts features from the historical hydrological monitoring data and riverbed evolution data of each target river channel to determine the river hydrological feature vector and river evolution vector of each target river channel; Acquisition Unit: Acquire multiple water level stations in the target waterway section, as well as the water level station number, coordinates, and data of each water level station. Extract features from the water level station data of each water level station to determine the water level station feature vector of each water level station. River water level station set unit: Based on the water level station feature vectors and coordinates of all water level stations, as well as the river hydrological feature vector and river center coordinates of each target river, the river water level station set of each target river is determined. The river water level station set includes multiple water level station numbers. River hydrological prediction model unit: Based on the river center coordinates of all target rivers of each river type, historical hydrological monitoring data, and the coordinates of each water level station number in the river water level station set, a river hydrological prediction model is constructed for each river type.

[0007] Preferably, a graphics-accelerated electronic navigation chart navigation element rendering system includes a sequence module comprising: Monitoring Unit: Determines the real-time monitoring time based on the monitoring frequency, acquires real-time water level monitoring data for each target river channel based on the real-time monitoring time, and preprocesses the real-time water level monitoring data for each target river channel; Initial time series unit: The next monitoring time is obtained based on the monitoring frequency and real-time monitoring time; the prediction time interval is determined based on the river type of each target river; and the initial time series of each target river is determined based on the real-time monitoring time, the next monitoring time, and the prediction time interval of each target river. Initial prediction reference data unit: The river center coordinates, river water level station coordinates, real-time water level monitoring data and initial time series of each target river are input into the river hydrological prediction model of the corresponding river type to determine the initial prediction reference data of each target river. The initial prediction reference data includes the prediction reference surface of each prediction time point in the initial time series. Initial dynamic water depth data unit: Extract static mapping water depth data for each target channel from the electronic waterway chart. Based on the static mapping water depth data, gridded waterway model, and river sedimentation simulation model for each target channel, determine the initial dynamic water depth data for each target channel. The initial dynamic water depth data includes the dynamic water depth field at each prediction time point in the initial time series.

[0008] Preferably, the sequence module of the graphics-accelerated electronic navigation chart navigation element rendering system further includes: Vessel navigation data unit: Acquires vessel navigation data for the target channel section, including standard vessel draft, safe margin depth, and dangerous margin depth; Status label unit: Based on the vessel navigation data of the target channel section and the dynamic water depth field of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel, determine the status label of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel. The status labels include safety, caution, and danger. Write-back time series unit: Based on the status labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, determine the write-back time series of each target channel.

[0009] Preferably, a graphics-accelerated electronic navigation chart navigation element rendering system, writing back to time-series units, includes: Initial state sequence sub-unit: Based on the state labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, the initial state sequence of each target channel is determined. Judgment sub-unit: If two adjacent state labels are the same in the initial state sequence of the target river channel, determine the predicted time point corresponding to the first state label in the initial time sequence as the adjustment time point, and discard the predicted time point corresponding to the second state label in the initial time sequence. Adjustment time point subunit: If the status labels of every two adjacent predicted time points in the initial time series of the initial dynamic water depth data of the target river channel are different, the two adjacent predicted time points are determined as adjustment time points respectively; Interpolation time point subunit: Based on the state labels of two adjacent prediction time points with different state labels in the initial time series of the initial dynamic water depth data of the target river channel and the dynamic water depth field, the interpolation time point of the two adjacent prediction time points is determined; Iterative sub-unit: Iterate over all adjustment time points and all interpolation time points for each target river channel until two adjacent adjustment time points or interpolation time points satisfy the first rule or the second rule. The first rule is that the state labels are the same, and the second rule is that when the state labels are different, the time interval between two adjacent adjustment time points or interpolation time points is less than or equal to the preset minimum time interval. Write-back time series sub-unit: Based on all adjustment time points and all interpolation time points of the last iteration for each target channel, determine the write-back time series for each target channel.

[0010] Preferably, the write-back module of the graphics-accelerated electronic navigation chart navigation element rendering system includes: Electronic navigation data unit: Acquires electronic navigation data based on standard data format from electronic navigation charts. The electronic navigation data includes at least depth point storage data and depth derived data, and acquires depth value encoding rules. The depth derived data includes multiple original derived elements. Predicted water depth value unit: Based on the coordinates of each stored water depth point in the water depth point storage data, spatial interpolation is performed on the dynamic water depth field at each adjustment time point or interpolation time point in the write-back time series to determine the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series. Encoded water depth value unit: Based on the water depth value encoding rules, the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series is encoded to determine the encoded water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series; Depth Point Replacement Data Unit: Locates the record identifier of each stored depth point in the depth point storage data, writes the coded depth value of each stored depth point at each adjustment time point or interpolation time point in the write-back time series back to the record position of the corresponding stored depth point, and determines the depth point replacement data based on the standard data format for each adjustment time point or interpolation time point in the write-back time series. Derivative Replacement Data Unit: Based on the water depth point replacement data of each adjustment time point or interpolation time point in the write-back time series, generate derivative replacement data of each adjustment time point or interpolation time point in the write-back time series based on the standard data format. The derivative replacement data includes multiple replacement derivative elements. Update channel data unit: Update electronic channel data based on standard data format depth point replacement data and derived replacement data for each adjustment or interpolation time point in the write-back time series, and determine the updated channel data for each adjustment or interpolation time point in the write-back time series.

[0011] Preferably, a graphics-accelerated electronic navigation chart navigation element rendering system includes a rendering module comprising: Rendering Method Unit: Obtains a set of derived feature types and a graphics-accelerated rendering method for each feature type in the set of derived feature types. The set of derived feature types includes at least point cloud features, point symbol features, line features, and complex building features. Updated Channel Chart Unit: A rendering method based on the feature type of each replacement derivative element in the derivative replacement data of the updated channel data at each adjustment time point or interpolation time point in the write-back time series is used to render the replacement derivative elements and determine the updated channel chart at each adjustment time point or interpolation time point in the write-back time series.

[0012] This invention provides a graphics-accelerated electronic navigation chart navigation element rendering platform for executing any of the graphics-accelerated electronic navigation chart navigation element rendering systems in Examples 1 to 7.

[0013] The beneficial effects of this invention compared to the prior art are as follows: the target waterway section is divided into multiple target channels according to a set interval; target channel data for each target channel is acquired and analyzed; multiple channel hydrological prediction models for channel types are determined; real-time water level monitoring data for each target channel is acquired; the dynamic water depth field and status label for each prediction time point in the initial time series of the target channel are determined; the write-back time series for each target channel is determined; electronic channel data based on standard data format of the electronic channel chart is acquired; updated channel data for each adjustment time point or interpolation time point in the write-back time series is determined; the updated channel data for each adjustment time point or interpolation time point in the write-back time series is rendered using graphics acceleration; and the updated channel chart for each adjustment time point or interpolation time point in the write-back time series is determined. It can solve the problems of incompatibility between dynamic data and standard formats, lag in static updates of electronic navigation charts, and low rendering efficiency. It realizes a closed-loop system with a full link, from segmented hydrological modeling of waterways, time-series dynamic water depth prediction, standardized data write-back to accelerated rendering of all elements of graphics, improving the real-time performance, accuracy and smooth operation of electronic navigation charts, and providing precise support for the entire process of inland waterway navigation safety.

[0014] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a graphics-accelerated electronic navigation chart navigation element rendering system according to an embodiment of the present invention. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Example 1:

[0018] This invention provides a graphics-accelerated electronic navigation chart navigation element rendering system, with reference to... Figure 1 ,include: Acquisition module: Divides the target waterway section into multiple target channels according to a set interval, acquires and analyzes the target channel data of each target channel, and determines the channel hydrological prediction model for multiple channel types; Sequence Module: Acquires real-time water level monitoring data for each target river channel. Based on the target river channel data, real-time water level monitoring data, and river hydrological prediction models for all river types, it determines the dynamic water depth field and status label for each prediction time point in the initial time series of the target river channel, and determines the write-back time series for each target river channel. Write-back module: Acquires electronic channel data based on standard data format from electronic channel charts, writes back the dynamic depth field of each adjustment time point or interpolation time point in the write-back time series based on standard data format, and determines the updated channel data for each adjustment time point or interpolation time point in the write-back time series; Rendering module: Performs graphics-accelerated rendering of the updated channel data for each adjustment or interpolation time point in the write-back time series, and determines the updated channel map for each adjustment or interpolation time point in the write-back time series.

[0019] In this embodiment, the acquisition module realizes the construction of refined channel segmentation and hydrological prediction model. First, the overall navigable range of the target channel section is determined. Based on the hydrological change law of inland waterway, a fixed segmentation interval is preset to divide the complete channel section into multiple continuous target river units with relatively uniform hydrological characteristics. Then, through the basic channel database, historical survey data and hydrological station records, the river type, center coordinates, historical hydrological monitoring data and riverbed evolution data of each target river are obtained to form complete target river data. Feature analysis and classification are performed on all target river data to classify multiple river types. Then, combined with the spatial coordinates, historical hydrological data and corresponding bound water level station information of the same type of river, a distributed parallel computing framework and machine learning algorithm are used to train and construct a river hydrological prediction model exclusive to each river type, providing core model support for the subsequent accurate calculation of dynamic water depth field.

[0020] In this embodiment, the sequence module realizes dynamic water depth time series prediction and write-back time series optimization. First, real-time water level monitoring data is collected from the water level stations bound to each target river at a preset monitoring frequency. The real-time data is cleaned, corrected, and preprocessed. Then, the river hydrological prediction model corresponding to the type of the target river is matched, and the real-time water level data and the basic data of the target river are synchronously input into the model. The start and end nodes and prediction interval of the initial time series are determined by combining the preset monitoring frequency. The model calculates and outputs the dynamic water depth field of the entire river at each prediction time point in the initial time series. Then, combined with the draft and margin parameters of navigation vessels in the channel, the dynamic water depth field of each prediction time point is classified for navigation safety, and three types of status labels of safety, attention, and danger are generated. Finally, based on the status labels of the entire time series, redundant time points with the same risk level are removed, key nodes of risk mutation are retained, and the transition interval is interpolated and optimized. After multi-rule iterative convergence, the write-back time series adapted to the navigation risk changes of each target river is determined.

[0021] In this embodiment, the write-back module realizes the standardized write-back of dynamic depth data and the update of channel data. First, it retrieves the original electronic channel data conforming to the CJ-57 standard data format from the electronic channel chart system, splits and extracts the depth point storage data and the depth derivative data, and obtains the depth value encoding rules corresponding to the standard format. Then, it traverses each adjustment time point and interpolation time point in the write-back time series, and performs spatial interpolation on the dynamic depth field of the corresponding time point based on the coordinates of the original depth points to obtain the predicted depth value of each original depth point. After completing the depth value encoding according to the standard encoding rules, it accurately locates the storage location in the original file through the unique record identifier of each depth point, completes the lossless write-back of the encoded depth value, and generates depth point replacement data. Then, based on the updated depth point data, it recalculates the isobaths, deep water area boundaries and other derivative elements to generate derivative replacement data conforming to the standard format. Finally, it uses the two types of replacement data to complete the full update of the original electronic channel data and generate the updated channel data corresponding to each time point.

[0022] In this embodiment, the rendering module achieves accelerated graphics rendering of all elements of the waterway map and generates updated waterway maps. First, it pre-constructs a set of waterway derivative element types, categorizing them into four core types based on their geometric characteristics and rendering requirements: water depth point clouds, point navigational aid symbols, linear waterway elements, and complex structures such as wharves and bridges. A dedicated GPU-accelerated rendering method is designed for each type of element. Then, it iterates through each adjustment and interpolation time point in the time series, retrieves the updated waterway data for the corresponding time point, extracts the type and attribute data of each replacement derivative element, matches the corresponding dedicated accelerated graphics rendering method, completes hardware-accelerated rendering of all elements, overlays a water depth field color gradation rendering layer and a waterway base map, and finally generates the updated waterway map for each time point. This achieves smooth and efficient rendering of high-density waterway elements in large scenes, ensuring the stable operation of the ship navigation terminal.

[0023] The beneficial effects of the above technologies are as follows: the target waterway section is divided into multiple target channels according to a set interval; target channel data for each target channel is acquired and analyzed; multiple channel hydrological prediction models for channel types are determined; real-time water level monitoring data for each target channel is acquired; the dynamic water depth field and status label for each prediction time point in the initial time series of the target channel are determined; the write-back time series for each target channel is determined; electronic channel data based on standard data format of the electronic channel chart is acquired; updated channel data for each adjustment or interpolation time point in the write-back time series is determined; the updated channel data for each adjustment or interpolation time point in the write-back time series is rendered using graphics acceleration; and the updated channel chart for each adjustment or interpolation time point in the write-back time series is determined. It can solve the problems of incompatibility between dynamic data and standard formats, lag in static updates of electronic navigation charts, and low rendering efficiency. It realizes a closed-loop system with a full link, from segmented hydrological modeling of waterways, time-series dynamic water depth prediction, standardized data write-back to accelerated rendering of all elements of graphics, improving the real-time performance, accuracy and smooth operation of electronic navigation charts, and providing precise support for the entire process of inland waterway navigation safety. Example 2:

[0024] Based on Example 1, the electronic navigation chart navigation element rendering system based on graphics acceleration includes an acquisition module comprising: Division Unit: The target waterway section is divided into multiple target waterways according to a set interval, and the target waterway data of each target waterway is obtained. The target waterway data includes waterway type, waterway center coordinates, historical hydrological monitoring data, and riverbed evolution data. Feature extraction unit: extracts features from the historical hydrological monitoring data and riverbed evolution data of each target river channel to determine the river hydrological feature vector and river evolution vector of each target river channel; Acquisition Unit: Acquire multiple water level stations in the target waterway section, as well as the water level station number, coordinates, and data of each water level station. Extract features from the water level station data of each water level station to determine the water level station feature vector of each water level station. River water level station set unit: Based on the water level station feature vectors and coordinates of all water level stations, as well as the river hydrological feature vector and river center coordinates of each target river, the river water level station set of each target river is determined. The river water level station set includes multiple water level station numbers. River hydrological prediction model unit: Based on the river center coordinates of all target rivers of each river type, historical hydrological monitoring data, and the coordinates of each water level station number in the river water level station set, a river hydrological prediction model is constructed for each river type.

[0025] In this embodiment, the target waterway segment is the entire range of a designated inland waterway requiring hydrological modeling and dynamic updates of electronic navigation charts. The set interval is a preset segment length threshold adapted to the changing patterns of the waterway's hydrological characteristics. The target river is a continuous sub-waterway unit with relatively uniform hydrological characteristics obtained after segmenting the target waterway segment. The target river data is the basic attributes and historical evolution dataset corresponding to each target river, including river type, river center coordinates, historical hydrological monitoring data, and riverbed evolution data. The river type can be a mountain river, a plain river, a tidal section, etc. The overall spatial boundary and navigation management scope of the target waterway segment are clearly defined. Combining the hydrological change characteristics and riverbed evolution patterns of the inland waterway, the complete target waterway segment is divided into multiple continuous target river units according to the set interval. Then, through the waterway basic database, historical multibeam scanning data, and historical hydrological station records, the river type, river center coordinates, historical hydrological monitoring data, and riverbed evolution data corresponding to each target river are obtained and integrated to form the target river data for each target river.

[0026] In this embodiment, the river hydrological feature vector is a standardized feature set obtained by extracting features from historical hydrological monitoring data of the target river. It is used to quantitatively characterize core hydrological features such as the river's water level change cycle, flow fluctuation pattern, and water depth evolution trend. The river evolution vector is a standardized feature set obtained by extracting features from the target river's riverbed evolution data. It is used to quantitatively characterize core evolutionary features such as the river's scouring and deposition rate, riverbed elevation fluctuation, and topographic evolution direction. Historical hydrological monitoring data and riverbed evolution data corresponding to each target river channel are retrieved. Data cleaning and standardization preprocessing are performed on the two types of data respectively. Outliers and systematic errors are automatically identified and removed. Then, through preset feature extraction rules, core hydrological attributes such as water level extremes during wet and dry seasons, water level change rate, and flow-water level correlation characteristics are extracted from the historical hydrological monitoring data to generate a river hydrological feature vector for each target river channel. At the same time, core topographic attributes such as interannual scour and deposition amplitude, shoal evolution trend, and riverbed topographic change characteristics are extracted from the riverbed evolution data to generate a river evolution vector for each target river channel, thus completing the quantitative representation of the hydrology and evolution patterns of the target river channels.

[0027] In this embodiment, a water level station is a fixed hydrological monitoring station deployed along the waterway for continuously collecting water level elevation data. The water level station number is a unique identification code used to distinguish each water level station. The water level station coordinates are the standardized spatial geographic coordinates of the station's location. The water level station data is a complete monitoring dataset including historical and real-time water level monitoring data, equipment operating status data, and data collection frequency. The water level station feature vector is a standardized set of features obtained after feature extraction from the water level station data, used to quantitatively characterize the station's core attributes such as water level change patterns, data continuity, and monitoring accuracy. The list of all compliant water level stations within the target waterway section is obtained through the waterway hydrological monitoring network. Simultaneously, the unique water level station number, spatial geographic coordinates, and full water level station data for each station are retrieved. Then, the water level station data for each station is cleaned and outliers are removed to ensure the quality and reliability of the input data. Then, through preset feature extraction rules, core features such as water level time series change characteristics, data completeness, water level fluctuation amplitude during wet and dry seasons, and water level response rate are extracted from the water level station data to generate a standardized water level station feature vector for each station, thus completing the quantitative representation of the hydrological patterns of the monitoring stations.

[0028] In this embodiment, the river level station set is a collection of numbered stations that are highly correlated with the hydrological characteristics of the corresponding target river and spatially matched. It is a dedicated monitoring station group that provides core data support for the hydrological prediction of the target river. The process involves retrieving the feature vectors and coordinates of all level stations, as well as the hydrological feature vector and center coordinates of each target river. First, the linear spatial distance between each level station and the center coordinates of each target river is calculated based on the spatial geographic coordinate system. Level stations with satisfactory spatial proximity are selected for initial screening. Then, based on the feature vectors of the level stations and the hydrological feature vectors of the target rivers, the similarity of their hydrological change patterns is calculated to complete a secondary verification of the hydrological feature correlation. Combining the calculation results of spatial proximity and hydrological feature similarity, one or more level stations that are highly matched with the corresponding target river are finally selected. The unique numbers of these level stations are extracted and integrated to form the river level station set corresponding to each target river, establishing a precise binding relationship between the target river and the monitoring stations.

[0029] In this embodiment, the river hydrological prediction model is constructed for different river types. It is a hydrodynamic and machine learning fusion model used to predict the water depth changes and hydrological dynamics of target rivers, and is the core computing carrier for generating dynamic water depth fields of electronic navigation charts. All target rivers are classified and collected according to their river types. All target rivers of the same type are integrated into a dedicated model training dataset. Then, the river center coordinates, historical hydrological monitoring data, and the coordinates and historical monitoring data of each water level station corresponding to the water level station number in the set of river water level stations for each target river are retrieved from each dataset. The above multi-source data are spatiotemporally registered and aligned in a unified spatiotemporal coordinate system to eliminate systematic errors in the spatiotemporal dimension. Then, using the historical monitoring data of the water level stations as input and the actual hydrological change data of the corresponding target river as output, the training, validation, and hyperparameter optimization of the model are completed using an appropriate machine learning algorithm or hydrodynamic model. Finally, a dedicated river hydrological prediction model corresponding to each river type is constructed.

[0030] In this embodiment, the river hydrological prediction model is based on the similarity and spatial correlation of hydrological characteristics of different types of inland rivers. It constructs a categorized water level-depth spatiotemporal response prediction model. First, it uses the coordinates of the river center and the coordinates of the water level station as spatial anchor points to establish a spatial mapping relationship between the entire target river area and the monitoring data of the corresponding water level station. Then, it uses long-term historical hydrological monitoring data of all target rivers of the same type as the training basis to fit the inherent correlation between river water level changes and the overall water depth distribution under different hydrological conditions. The model training convergence and generalization optimization are completed through machine learning algorithms or hydrodynamic equations. Finally, it can accurately predict the hydrological dynamic change process of the entire target river area based on real-time water level station monitoring data, providing core model support for dynamic water depth field calculation.

[0031] The beneficial effects of the above technologies are as follows: the target waterway section is divided into multiple target waterways according to a set interval, the target waterway data of each target waterway is acquired and analyzed, and multiple waterway hydrological prediction models are determined. This can adapt to complex hydrological changes in waterways, improve the accuracy and scenario adaptability of hydrological prediction, and provide reliable model support for the dynamic updating of electronic waterway charts. Example 3:

[0032] Based on Example 2, the electronic navigation chart navigation element rendering system based on graphics acceleration includes a sequence module, comprising: Monitoring Unit: Determines the real-time monitoring time based on the monitoring frequency, acquires real-time water level monitoring data for each target river channel based on the real-time monitoring time, and preprocesses the real-time water level monitoring data for each target river channel; Initial time series unit: The next monitoring time is obtained based on the monitoring frequency and real-time monitoring time; the prediction time interval is determined based on the river type of each target river; and the initial time series of each target river is determined based on the real-time monitoring time, the next monitoring time, and the prediction time interval of each target river. Initial prediction reference data unit: The river center coordinates, river water level station coordinates, real-time water level monitoring data and initial time series of each target river are input into the river hydrological prediction model of the corresponding river type to determine the initial prediction reference data of each target river. The initial prediction reference data includes the prediction reference surface of each prediction time point in the initial time series. Initial dynamic water depth data unit: Extract static mapping water depth data for each target channel from the electronic waterway chart. Based on the static mapping water depth data, gridded waterway model, and river sedimentation simulation model for each target channel, determine the initial dynamic water depth data for each target channel. The initial dynamic water depth data includes the dynamic water depth field at each prediction time point in the initial time series.

[0033] In this embodiment, the monitoring frequency is a preset water level data collection interval adapted to the river's hydrological changes, such as 1 hour. The real-time monitoring time is a precise time node determined based on the monitoring frequency to trigger water level data collection. Real-time water level monitoring data is the real-time water level elevation data collected from the water level stations bound to the target river at the corresponding time point. Preprocessing is a standardized process for cleaning and correcting the collected data. First, based on the severity of river hydrological changes and navigation safety requirements, a suitable water level monitoring frequency is preset. A fixed real-time monitoring time node is determined based on the time interval of the monitoring frequency. When this time node is reached, real-time water level monitoring data for the target river is synchronously collected from all water level stations in the set of river water level stations corresponding to each target river. Then, the collected real-time water level monitoring data is preprocessed using a preset data cleaning and anomaly detection algorithm library to automatically identify and remove abnormal data, thus completing the benchmark deviation correction.

[0034] In this embodiment, the next monitoring time is the next data acquisition time node based on the monitoring frequency. The prediction time interval is the time step between two adjacent prediction time points determined based on the river type of the target river. The initial time series is a standardized time series covering the complete prediction period from the real-time monitoring time to the next monitoring time, containing multiple consecutive prediction time points. Based on the preset monitoring frequency, starting from the current real-time monitoring time, the next monitoring time node is determined by extending the time interval corresponding to one monitoring frequency. Then, according to the river type of each target river, the corresponding river's hydrological change rate and scour-deposition evolution law are matched to determine the appropriate prediction time interval. The more drastic the river's hydrological changes, the shorter the prediction time interval is set. For example, 10 minutes for mountain rivers and 20 minutes for plain rivers. Then, starting from the real-time monitoring time and ending at the next monitoring time, uniform interpolation is performed according to the determined prediction time interval to generate a complete time series containing multiple consecutive prediction time points. Finally, the initial time series corresponding to each target river is determined.

[0035] In this embodiment, the prediction reference surface is a water depth reference elevation surface based on the dynamic changes of real-time water level. The initial prediction reference data is a full-time reference dataset containing the prediction reference surface at each prediction time point in the initial time series. The coordinates of the river center corresponding to each target river, the coordinates of all water level stations in the river water level station set, the preprocessed real-time water level monitoring data, and the initial time series corresponding to the target river are retrieved. Then, the data is matched with the dedicated river hydrological prediction model corresponding to the river type to which the target river belongs. All the above data are synchronously input into the matched river hydrological prediction model. The model completes the full-time hydrological extrapolation calculation through a distributed parallel computing framework, outputting the dynamic elevation reference corresponding to each prediction time point in the initial time series, which is the prediction reference surface. Finally, the prediction reference surfaces of all time points are integrated and aggregated to determine the initial prediction reference data corresponding to each target river.

[0036] In this embodiment, the static mapping water depth data is the static mapping elevation and water depth data of the underwater topography of the waterway extracted from the original electronic waterway chart file conforming to the CJ-57 standard. The gridded waterway model is an irregular triangular network or regular grid model constructed by gridding the underwater topography of the target river channel. The river channel sedimentation projection model is a machine learning or hydrodynamic model trained based on historical hydrological and topographic data, used to predict the short-term scouring and sedimentation evolution trend of the riverbed. The initial dynamic water depth data is a dynamic water depth field dataset containing the spatial water depth distribution of the entire river channel at each prediction time point in the initial time series. From the original electronic navigation chart files conforming to the CJ-57 standard, the full-range static mapping water depth data corresponding to each target river is extracted to construct the static digital water depth base of the target river. Based on the gridded differential calculation of the unified static digital water depth base, a water depth spatial distribution dataset covering the entire target river area and time series is generated. First, the riverbed elevation of each node of the gridded navigation model after river sedimentation simulation model correction is used as the topographic base. Combined with the dynamic water level datum data corresponding to each prediction time point in the initial time series, the distributed parallel computing framework is called to perform standardized differential water depth calculation on each grid node. Then, all prediction time points in the initial time series are traversed to complete the water depth calculation of all grid nodes of the entire river, forming a water depth spatial distribution matrix covering the entire target river area at the corresponding time point, that is, the dynamic water depth field of a single time point. Finally, the dynamic water depth fields of all prediction time points are time-series collected to form the initial dynamic water depth data of the target river, providing a core data source for subsequent electronic navigation chart data updates and navigation safety assessments.

[0037] In this embodiment, the gridded waterway model is based on discrete grid partitioning technology, which transforms the continuous underwater topographic space of the target river into a standardized spatial carrier that can be finely parallelized. First, based on the static mapping water depth data extracted from the electronic waterway chart, and combined with the river boundary, center coordinates and underwater topographic feature point data of the target river, the full-element spatial registration is completed under a unified national spatial coordinate system. Then, according to the river width, topographic complexity and navigation importance, an appropriate grid partitioning mode is selected. The grid precision is increased in shallow shoal and bend areas with drastic topographic changes, while the grid precision is appropriately relaxed in straight waterway areas with gentle topography. The continuous underwater topography of the river is discretized into a series of interconnected grid units. Each grid node is matched with the corresponding static mapping water depth value and spatial coordinates. At the same time, the spatial index and topological association relationship of the grid units are established, and finally a gridded waterway model covering the entire target river area is formed, providing a unified spatial computing basis for subsequent distributed parallel computing of the entire river depth.

[0038] In this embodiment, the riverbed sedimentation projection model is based on the dynamic laws of riverbed scouring and sedimentation evolution and the time-series prediction capability driven by historical data. It constructs a correction model for short-term changes in riverbed topography. First, it collects long-term historical water level data, sediment content data, and multibeam scanning riverbed topography data of the target river to build a training dataset. Through machine learning algorithms such as hydrodynamic models or long short-term memory networks, it fits the inherent correlation between river inflow and sediment conditions, hydrological cycle, riverbed scouring and sedimentation amplitude, and topographic evolution trend to complete the training and verification of the model. During real-time calculation, the model can combine the current hydrological conditions and time cycle to predict the riverbed elevation change of each grid node in the gridded navigation channel model of the target river, dynamically correct the static surveyed water depth data, and finally output accurate water depth calculation results that fit the actual underwater topography.

[0039] The beneficial effects of the above technologies are as follows: real-time water level monitoring data of each target river channel can be obtained. Based on the target river channel data, real-time water level monitoring data, and river hydrological prediction models of all river types, the dynamic water depth field of each prediction time point in the initial time series of the target river channel can be determined. This can achieve accurate matching between the prediction cycle and the monitoring cycle. The static mapping data and river sedimentation inference model can be linked to output a full-time dynamic water depth field, thereby improving the real-time performance and accuracy of the dynamic update of the electronic navigation chart. Example 4:

[0040] Based on Example 1, the sequence module of the graphics-accelerated electronic navigation chart navigation element rendering system further includes: Vessel navigation data unit: Acquires vessel navigation data for the target channel section, including standard vessel draft, safe margin depth, and dangerous margin depth; Status label unit: Based on the vessel navigation data of the target channel section and the dynamic water depth field of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel, determine the status label of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel. The status labels include safety, caution, and danger. Write-back time series unit: Based on the status labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, determine the write-back time series of each target channel.

[0041] In this embodiment, the vessel navigation data for the target channel section is the core set of boundary parameters ensuring safe navigation of vessels within the channel. The standard vessel draft is the full-load reference draft for vessels designed for navigation in the corresponding channel grade. The safety margin depth is the minimum excess water depth threshold to ensure normal navigation without the risk of bottoming out. The danger margin depth is the critical excess water depth threshold to indicate that a vessel faces the risk of bottoming out. First, the standard vessel draft parameters for the target channel section are determined by combining the channel grade, the type of vessels designed for navigation, and inland waterway navigation management regulations. Then, based on the channel bottom type, current velocity characteristics, and vessel maneuverability requirements, appropriate safety margin depths and danger margin depths are set. Finally, the three core parameters are integrated through the channel navigation management system to form complete vessel navigation data.

[0042] In this embodiment, based on the vessel navigation data of the target channel section and the dynamic water depth field of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel, the state label of each predicted time point in the initial dynamic water depth data of each target channel is determined. The calculation formula is expressed as follows: ; in, This represents the state label at the j-th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel. This represents the dynamic water depth field at the j-th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel, based on coordinates. The water depth, This represents the spatial region of the i-th target river channel. Let represent the minimum water depth of the dynamic water depth field at the j-th prediction time point in the initial time series of the initial dynamic water depth data of the i-th target channel, T represent the standard ship draft in the navigation data, SD represent the safe margin depth in the navigation data, and DD represent the dangerous margin depth in the navigation data.

[0043] The beneficial effects of the above technologies are as follows: determining the status label of each predicted time point in the initial time series of the target river and determining the write-back time series of each target river can solve the problems of uniform update frequency, delayed early warning during high-risk periods, and data redundancy during low-risk periods, thereby improving the pertinence and real-time performance of electronic waterway chart updates. Example 5:

[0044] Based on Example 4, the electronic navigation chart navigation element rendering system based on graphics acceleration writes back time series units, including: Initial state sequence sub-unit: Based on the state labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, the initial state sequence of each target channel is determined. Judgment sub-unit: If two adjacent state labels are the same in the initial state sequence of the target river channel, determine the predicted time point corresponding to the first state label in the initial time sequence as the adjustment time point, and discard the predicted time point corresponding to the second state label in the initial time sequence. Adjustment time point subunit: If the status labels of every two adjacent predicted time points in the initial time series of the initial dynamic water depth data of the target river channel are different, the two adjacent predicted time points are determined as adjustment time points respectively; Interpolation time point subunit: Based on the state labels of two adjacent prediction time points with different state labels in the initial time series of the initial dynamic water depth data of the target river channel and the dynamic water depth field, the interpolation time point of the two adjacent prediction time points is determined; Iterative sub-unit: Iterate over all adjustment time points and all interpolation time points for each target river channel until two adjacent adjustment time points or interpolation time points satisfy the first rule or the second rule. The first rule is that the state labels are the same, and the second rule is that when the state labels are different, the time interval between two adjacent adjustment time points or interpolation time points is less than or equal to the preset minimum time interval. Write-back time series sub-unit: Based on all adjustment time points and all interpolation time points of the last iteration for each target channel, determine the write-back time series for each target channel.

[0045] In this embodiment, the initial state sequence is a time-series label set formed by sequentially arranging the state labels corresponding to each predicted time point according to the chronological order of the initial time sequence. The initial dynamic water depth data of each target river channel is retrieved, and all predicted time points contained in the initial time sequence are extracted, along with the state labels corresponding to each predicted time point. Then, strictly following the chronological order of the initial time sequence, all state labels are sequentially arranged to form a time-series label sequence that perfectly matches the initial time sequence, ultimately determining the initial state sequence for each target river channel.

[0046] In this embodiment, the initial state sequence of each target river is traversed, and two adjacent state labels in the sequence are compared one by one. When two adjacent state labels are found to be completely identical, it means that the navigation safety risk level at these two time points has not changed, and there is no need to repeat the data write-back. At this time, the predicted time point corresponding to the first state label in the initial time sequence is determined as the adjustment time point for subsequent write-back, while the predicted time point corresponding to the second state label is discarded, thereby completing the initial elimination of redundant time points under the same risk level.

[0047] In this embodiment, the status labels corresponding to each pair of adjacent predicted time points in the sequence are compared one by one. When the status labels of two adjacent predicted time points are found to be different, it indicates that the navigation safety risk level between these two time points has changed and belongs to a key node that needs to be focused on and accurately updated. At this time, the two adjacent predicted time points are determined as the adjustment time points for subsequent write-back.

[0048] In this embodiment, based on the state labels of two adjacent prediction time points with different state labels in the initial time series of the initial dynamic water depth data of the target river channel, and the dynamic water depth field, the interpolation time point of the two adjacent prediction time points is determined. The calculation formula is expressed as follows: ; ; ; ; ; ; ; in, This represents the interpolation time point between the j-th and (j+1)-th predicted time points in the initial time series of the initial dynamic water depth data for the i-th target channel. This represents the interpolated sub-time point between the j-th and (j+1)-th predicted time points in the initial time series of the initial dynamic water depth data for the i-th target channel. This represents the state label at the (j+1)th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel. Let J and J represent the j-th and (j+1)-th prediction time points in the initial time series of the initial dynamic water depth data for the i-th target channel, respectively. Let represent the minimum water depth of the dynamic water depth field at the (j+1)th and (j-1)th prediction time points in the initial time series of the initial dynamic water depth data for the i-th target channel, respectively. This represents the curvature correction factor. Indicates the urgency level of the risk. This represents the safety and conservatism coefficient. This represents the weight of the state label at the j-th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel. This represents the target boundary at the j-th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel. Let represent the water depth change rate and water depth change curvature, respectively, of the dynamic water depth field at the j-th and (j+1)-th prediction time points in the initial time series of the initial dynamic water depth data for the i-th target channel. The minimum water depth of the dynamic water depth field at the j-th prediction time point in the initial time series of the initial dynamic water depth data for the i-th target channel is represented by the weighted distance based on the target boundary. Represents a symbolic function. This indicates the preset minimum time interval.

[0049] In this embodiment, the target boundary This refers to the critical water depth threshold for switching the corresponding status label.

[0050] In this embodiment, the sign function If x > 0, =1, if x < 0, =-1, if x=0, =0.

[0051] In this embodiment, Indicates the interpolation time point and the j-th prediction time point The time interval.

[0052] In this embodiment, the curvature correction coefficient The basic influence weight of the control acceleration on the time interval is set, with a value ranging from 0.5 to 1. The value is determined according to the degree of change in the target river channel. For example, a higher value is used for rapid currents or soft muddy riverbeds, and a lower value is used for stable river channels.

[0053] In this embodiment, the safety conservatism factor The overall scaling time interval is set to ensure early warning. The value range is 0.6-0.8, with 0.7-0.8 for regular waterways and 0.6-0.7 for restricted waterways such as shallows.

[0054] In this embodiment, the risk urgency factor Control when very close to the danger boundary, i.e. The additional lead time for smaller values ​​ranges from 0.1 to 0.3. Larger values ​​indicate that the interpolation point is closer to the boundary. To move closer.

[0055] In this embodiment, the preset minimum time interval is the minimum time step threshold that adapts to the update frequency of electronic navigation charts and navigation safety requirements. After the initial time point screening is completed, all adjustment and interpolation time points for each target river are first collected and arranged in chronological order. Then, adjacent time points are iteratively verified group by group. During the verification process, two preset rules are followed. The first rule is that when the status labels corresponding to two adjacent time points are the same, the iterative optimization of that group is stopped. The second rule is that when the status labels corresponding to two adjacent time points are different, the time interval between the two time points is verified. If the interval is less than or equal to the preset minimum time interval, the iterative optimization of that group is stopped. If the interval is greater than the preset minimum time interval, the interpolation optimization and status verification continue. This process is repeated until all adjacent time points satisfy the first rule or the second rule, thereby completing the refined iterative optimization of the write-back time series.

[0056] In this embodiment, after the iterative subunit completes the verification and optimization of all adjacent time points and obtains all the adjustment and interpolation time points that conform to the rules from the last iteration output of each target river, these time points are strictly sorted in chronological order, and duplicate or invalid time nodes are removed to form a continuous and standardized time series. This is ultimately determined as the write-back time series corresponding to each target river. This series not only eliminates redundant time points of the same risk level but also fully covers the key nodes and transition intervals of risk abrupt changes.

[0057] The beneficial effects of the above technologies are as follows: Based on the status labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target river, the write-back time series of each target river can be determined, which can improve the accuracy and operational efficiency of electronic navigation chart data updates, ensure the update granularity of high-risk areas, reduce the ineffective computing power consumption of low-risk areas, and fully adapt to the dynamic safety assurance requirements of inland waterway navigation. Example 6:

[0058] Based on Example 1, the write-back module of the graphics-accelerated electronic navigation chart navigation element rendering system includes: Electronic navigation data unit: Acquires electronic navigation data based on standard data format from electronic navigation charts. The electronic navigation data includes at least depth point storage data and depth derived data, and acquires depth value encoding rules. The depth derived data includes multiple original derived elements. Predicted water depth value unit: Based on the coordinates of each stored water depth point in the water depth point storage data, spatial interpolation is performed on the dynamic water depth field at each adjustment time point or interpolation time point in the write-back time series to determine the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series. Encoded water depth value unit: Based on the water depth value encoding rules, the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series is encoded to determine the encoded water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series; Depth Point Replacement Data Unit: Locates the record identifier of each stored depth point in the depth point storage data, writes the coded depth value of each stored depth point at each adjustment time point or interpolation time point in the write-back time series back to the record position of the corresponding stored depth point, and determines the depth point replacement data based on the standard data format for each adjustment time point or interpolation time point in the write-back time series. Derivative Replacement Data Unit: Based on the water depth point replacement data of each adjustment time point or interpolation time point in the write-back time series, generate derivative replacement data of each adjustment time point or interpolation time point in the write-back time series based on the standard data format. The derivative replacement data includes multiple replacement derivative elements. Update channel data unit: Update electronic channel data based on standard data format depth point replacement data and derived replacement data for each adjustment or interpolation time point in the write-back time series, and determine the updated channel data for each adjustment or interpolation time point in the write-back time series.

[0059] In this embodiment, the standard data format is either the CJ-57 standard commonly used for inland waterway electronic navigation charts or the internationally recognized S-57 standard data format. Electronic navigation data is the complete set of navigation element data from the original data source file of the electronic navigation chart, conforming to the standard format. Depth point storage data is a dataset of coordinates and depth values ​​for all statically mapped depth points stored in the file. Depth-derived data consists of original derived elements such as isobaths, deep-water area boundaries, and safety depth lines, calculated based on the original depth points. Depth value encoding rules are the storage and encoding specifications for depth values ​​defined in the standard format. From the original standard file of the electronic navigation chart system, the complete electronic navigation data conforming to the CJ-57 standard data format is retrieved. The depth point storage data and depth-derived data containing multiple original derived elements are extracted from the data. Simultaneously, the corresponding depth value encoding rules for this standard format are retrieved, completing the aggregation of the original data and encoding specifications.

[0060] In this embodiment, the spatial geographic coordinates corresponding to each original stored water depth point in the water depth point storage data are retrieved first. Then, the dynamic water depth field of the entire river channel corresponding to each adjustment time point or interpolation time point in the write-back time series is retrieved. Using the coordinates of each stored water depth point as the target position, spatial interpolation calculation is performed in the gridded model of the dynamic water depth field to accurately match the dynamic water depth value at the corresponding time point, i.e., the predicted water depth value. By traversing all stored water depth points and all time points in the write-back time series, the predicted water depth value corresponding to each stored water depth point at each time point is finally determined, so as to achieve accurate spatial matching between the dynamic prediction data and the original channel water depth points.

[0061] In this embodiment, the water depth value encoding rules corresponding to the standard format are retrieved, and then the predicted water depth value corresponding to each stored water depth point at each adjustment time point or interpolation time point in the time series is retrieved. The numerical precision storage format and encoding method specified in the encoding rules are strictly followed, and each predicted water depth value is standardized and encoded to generate an encoded water depth value that meets the requirements of the original file format. All stored water depth points at all time points are traversed, and finally the encoded water depth value corresponding to each stored water depth point at each time point is determined, ensuring that the encoding format of the dynamic water depth data is fully compatible with the original file.

[0062] In this embodiment, the record identifier is a unique identifier corresponding to each depth point in the standard data format, used to accurately locate the storage location of the depth point in the original file. First, the unique record identifier corresponding to each stored depth point in the depth point storage data is retrieved. The storage record location of each depth point in the original standard file is accurately located using the identifier. Then, the coded depth value corresponding to each stored depth point at each time point in the write-back time series is written point by point to the corresponding storage location in the original file, completing the dynamic replacement of the original static depth value. After each full depth point write-back operation at a time point is completed, depth point replacement data conforming to the standard data format at that time point is generated. By traversing all time points in the write-back time series, the depth point replacement data corresponding to each time point is finally determined, realizing the standardized and lossless replacement of the original static data by dynamic depth data, which is fully compatible with all standard-compliant upper-layer applications.

[0063] In this embodiment, the replacement data of water depth points corresponding to each adjustment or interpolation time point in the write-back time series is retrieved. Using the updated full water depth point data as the calculation basis, the associated element calculation module is automatically triggered to recalculate all derived channel elements that depend on the water depth data, including isobaths, deep-water area boundaries, safety depth lines, obstacles, bridges, and other navigation aids. Multiple replacement derived elements that completely match the updated water depth data are generated. Then, in strict accordance with the requirements of standard data format, all replacement derived elements are standardized, organized, and coded. Finally, the derivative replacement data corresponding to each time point that conforms to the standard data format is generated, realizing the linkage and synchronous update of water depth-related elements and ensuring the logical consistency of all elements in the channel chart.

[0064] In this embodiment, the original electronic navigation channel data conforming to the standard data format is retrieved, and then the depth point replacement data and derived replacement data corresponding to each adjustment time point or interpolation time point in the write-back time series are retrieved. The two types of replacement data are completely written into the corresponding layer and landmark position of the original electronic navigation channel data to complete the full dynamic update of the original data. After each time point completes the full element update operation, the complete navigation channel dataset corresponding to that time point is generated, i.e., the updated navigation channel data. By traversing all time points of the write-back time series, the updated navigation channel data corresponding to each time point is finally determined.

[0065] The beneficial effects of the above technologies are as follows: acquiring electronic navigation data based on standard data format from electronic navigation charts; writing back the dynamic water depth field at each adjustment or interpolation time point in the write-back time series based on the standard data format; determining the updated navigation data at each adjustment or interpolation time point in the write-back time series; achieving a fully standardized replacement of the original navigation data with dynamic water depth data; solving the problems of logical confusion and information inconsistency caused by overlay layers; achieving zero interference and zero adaptation for upper-layer applications; significantly improving the standardization and data consistency of dynamic updates of electronic navigation charts; and comprehensively ensuring the accuracy and compatibility of navigation data. Example 7:

[0066] Based on Example 1, the rendering module of the graphics-accelerated electronic navigation chart navigation element rendering system includes: Rendering Method Unit: Obtains a set of derived feature types and a graphics-accelerated rendering method for each feature type in the set of derived feature types. The set of derived feature types includes at least point cloud features, point symbol features, line features, and complex building features. Updated Channel Chart Unit: A rendering method based on the feature type of each replacement derivative element in the derivative replacement data of the updated channel data at each adjustment time point or interpolation time point in the write-back time series is used to render the replacement derivative elements and determine the updated channel chart at each adjustment time point or interpolation time point in the write-back time series.

[0067] In this embodiment, the derived element type set is a collection of all depth-derived elements and navigation elements in the electronic navigation chart, categorized according to their graphic rendering characteristics. It includes at least four core element types: point cloud elements, point symbol elements, line elements, and complex building elements. The graphics-accelerated rendering method utilizes a GPU hardware-accelerated rendering pipeline and dedicated optimization algorithms adapted to the graphic characteristics of different elements. All navigation-derived elements in the electronic navigation chart are classified and grouped according to their graphic features. Based on the geometric rendering requirements and business attributes of the elements, a derived element type set is formed, including point cloud elements, point symbol elements, line elements, and complex building elements. Then, a dedicated graphics-accelerated rendering method is designed to match the geometric characteristics and rendering challenges of each element type in the set.

[0068] In this embodiment, for point cloud elements, efficient parallel rendering of point cloud data is achieved based on octree spatial indexing and GPU instantiation rendering technology. First, the point cloud data of the entire waterway depth is pre-graded and processed to construct an octree spatial index structure. The point cloud data is organized into blocks according to spatial location and level of detail. During the rendering process, the frustum culling algorithm is used to filter out point cloud blocks outside the current camera's field of view, retaining only the valid data within the field of view. Then, the GPU instantiation rendering pipeline is used to share the vertex data of point clouds at the same level to the same vertex buffer. The GPU completes the coordinate transformation, color mapping and rendering operations of all point clouds in parallel. At the same time, combined with the hierarchical detail technology, the distant point clouds are downsampled and merged to reduce the total number of rendered vertices, which greatly reduces the CPU scheduling overhead and GPU rendering computing power consumption, and achieves smooth rendering of large-scene water depth point clouds.

[0069] In this embodiment, for point-like symbol elements, efficient rendering of high-density point-like elements is achieved based on GPU instantiation rendering and symbol aggregation technology. First, point-like symbols such as navigation marks and water level stations are classified and processed, and a shared symbol template mesh and style parameter set are constructed for symbols of the same type. During the rendering process, the GPU instantiation rendering pipeline expands the single drawing instructions of symbols of the same type into batch drawing. The GPU completes the coordinate positioning, style calculation and texture mapping of each symbol in parallel, without the need for the CPU to schedule the drawing of each symbol one by one. At the same time, combined with the spatial aggregation algorithm, in high-density flight segments or high scaling levels, symbols of the same type that are too close to each other in the field of view are automatically aggregated and merged into a single aggregated symbol for drawing, which greatly reduces the number of drawing calls and the total number of rendered primitives, effectively solving the problem of frame rate drop caused by high-density point-like symbol rendering.

[0070] In this embodiment, for linear features, smooth and efficient rendering of multi-segment lines is achieved based on GPU geometry shaders and frustum culling technology. First, linear features such as contour lines, channel boundaries, and report lines are pre-graded. Multi-level detail levels and spatial indexes are constructed according to the line segment precision and spatial range. During the rendering process, the frustum culling algorithm filters out line segments outside the field of view, retaining only the valid line segment data within the field of view. Then, the line segment vertex data is input into the GPU rendering pipeline, and the geometry shader performs edge expansion, corner smoothing, and anti-aliasing processing of the line segments in parallel. There is no need for the CPU to pre-process the geometric topology of the line segments. At the same time, combined with graded detail technology, the line segment data with corresponding precision is automatically matched at different scaling levels. The number of line segment vertices is automatically simplified in long-distance scenes, which greatly reduces the total amount of geometric primitives rendered, and achieves smooth and fluid rendering of long-distance, high-density linear features.

[0071] In this embodiment, for complex architectural elements, efficient rendering of complex models in large scenes is achieved based on model block streaming loading and GPU mesh instantiation rendering technology. First, the 3D models of complex buildings such as docks, bridges, and locks are preprocessed and divided into mesh blocks according to spatial location. Multi-level detail layers are constructed for each block model, and a lightweight compression algorithm is used to compress the model mesh and texture. During the rendering process, the view frustum prediction algorithm is combined to predict the area that the user's field of view will enter and request the model block data of the corresponding area in advance through streaming loading. Model blocks that leave the field of view are unloaded from the video memory in time. At the same time, through the GPU mesh instantiation rendering pipeline, the same mesh data is shared for repeated architectural components. The GPU completes the coordinate transformation, texture mapping, and lighting calculation of batch meshes in parallel, which greatly reduces the video memory usage and CPU scheduling overhead, avoids the video memory overflow problem caused by large-scale model rendering, and achieves smooth interaction and rendering of complex architectural scenes.

[0072] In this embodiment, the updated waterway data corresponding to each adjustment or interpolation time point in the write-back time series is retrieved. From the derived replacement data of the updated waterway data, the feature type and complete attribute data corresponding to each replacement derivative element are extracted. Then, the dedicated graphics-accelerated rendering method corresponding to the feature type in the rendering method unit is matched. The geometric data attribute data and style parameters of the replacement derivative element are input into the corresponding GPU rendering pipeline. Hardware-accelerated rendering of the corresponding element is completed through GPU parallel computing. After traversing all replacement derivative elements at that time point and completing the rendering of all elements, the waterway base map and water depth field color level rendering layer are superimposed to generate the complete visualized electronic waterway map result corresponding to that time point, i.e., the updated waterway map. By traversing all time points in the write-back time series, the updated waterway map corresponding to each time point is finally determined, realizing efficient and smooth rendering of all elements of the electronic waterway map.

[0073] The beneficial effects of the above technologies are as follows: rendering the updated waterway data at each adjustment or interpolation time point in the write-back time series based on graphics acceleration, and determining the updated waterway map at each adjustment or interpolation time point in the write-back time series, can improve the rendering efficiency and image accuracy of large-scene waterway maps, while taking into account the visualization accuracy and operational smoothness of navigation elements, and providing stable and accurate real-time visualization support for inland waterway vessel navigation. Example 8:

[0074] This invention provides a graphics-accelerated electronic navigation chart navigation element rendering platform for executing any of the graphics-accelerated electronic navigation chart navigation element rendering systems in Examples 1 to 7.

[0075] The beneficial effects of the above technologies are as follows: the target waterway section is divided into multiple target channels according to a set interval; target channel data for each target channel is acquired and analyzed; multiple channel hydrological prediction models for channel types are determined; real-time water level monitoring data for each target channel is acquired; the dynamic water depth field and status label for each prediction time point in the initial time series of the target channel are determined; the write-back time series for each target channel is determined; electronic channel data based on standard data format of the electronic channel chart is acquired; updated channel data for each adjustment or interpolation time point in the write-back time series is determined; the updated channel data for each adjustment or interpolation time point in the write-back time series is rendered using graphics acceleration; and the updated channel chart for each adjustment or interpolation time point in the write-back time series is determined. It can solve the problems of incompatibility between dynamic data and standard formats, lag in static updates of electronic navigation charts, and low rendering efficiency. It realizes a closed-loop system with a full link, from segmented hydrological modeling of waterways, time-series dynamic water depth prediction, standardized data write-back to accelerated rendering of all elements of graphics, improving the real-time performance, accuracy and smooth operation of electronic navigation charts, and providing precise support for the entire process of inland waterway navigation safety.

[0076] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A graphics-accelerated electronic navigation chart navigation element rendering system, characterized in that, include: Acquisition module: Divides the target waterway section into multiple target channels according to a set interval, acquires and analyzes the target channel data of each target channel, and determines the channel hydrological prediction model for multiple channel types; Sequence Module: Acquires real-time water level monitoring data for each target river channel. Based on the target river channel data, real-time water level monitoring data, and river hydrological prediction models for all river types, it determines the dynamic water depth field and status label for each prediction time point in the initial time series of the target river channel, and determines the write-back time series for each target river channel. Write-back module: Acquires electronic channel data based on standard data format from electronic channel charts, writes back the dynamic depth field of each adjustment time point or interpolation time point in the write-back time series based on standard data format, and determines the updated channel data for each adjustment time point or interpolation time point in the write-back time series; Rendering module: Performs graphics-accelerated rendering of the updated channel data for each adjustment or interpolation time point in the write-back time series, and determines the updated channel map for each adjustment or interpolation time point in the write-back time series.

2. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 1, characterized in that, The acquisition module includes: Division Unit: The target waterway section is divided into multiple target waterways according to a set interval, and the target waterway data of each target waterway is obtained. The target waterway data includes waterway type, waterway center coordinates, historical hydrological monitoring data, and riverbed evolution data. Feature extraction unit: extracts features from the historical hydrological monitoring data and riverbed evolution data of each target river channel to determine the river hydrological feature vector and river evolution vector of each target river channel; Acquisition Unit: Acquire multiple water level stations in the target waterway section, as well as the water level station number, coordinates, and data of each water level station. Extract features from the water level station data of each water level station to determine the water level station feature vector of each water level station. River water level station set unit: Based on the water level station feature vectors and coordinates of all water level stations, as well as the river hydrological feature vector and river center coordinates of each target river, the river water level station set of each target river is determined. The river water level station set includes multiple water level station numbers. River hydrological prediction model unit: Based on the river center coordinates of all target rivers of each river type, historical hydrological monitoring data, and the coordinates of each water level station number in the river water level station set, a river hydrological prediction model is constructed for each river type.

3. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 2, characterized in that, Sequence module, including: Monitoring Unit: Determines the real-time monitoring time based on the monitoring frequency, acquires real-time water level monitoring data for each target river channel based on the real-time monitoring time, and preprocesses the real-time water level monitoring data for each target river channel; Initial time series unit: The next monitoring time is obtained based on the monitoring frequency and real-time monitoring time; the prediction time interval is determined based on the river type of each target river; and the initial time series of each target river is determined based on the real-time monitoring time, the next monitoring time, and the prediction time interval of each target river. Initial prediction reference data unit: The river center coordinates, river water level station coordinates, real-time water level monitoring data and initial time series of each target river are input into the river hydrological prediction model of the corresponding river type to determine the initial prediction reference data of each target river. The initial prediction reference data includes the prediction reference surface of each prediction time point in the initial time series. Initial dynamic water depth data unit: Extract static mapping water depth data for each target channel from the electronic waterway chart. Based on the static mapping water depth data, gridded waterway model, and river sedimentation simulation model for each target channel, determine the initial dynamic water depth data for each target channel. The initial dynamic water depth data includes the dynamic water depth field at each prediction time point in the initial time series.

4. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 1, characterized in that, The sequence module also includes: Vessel navigation data unit: Acquires vessel navigation data for the target channel section, including standard vessel draft, safe margin depth, and dangerous margin depth; Status label unit: Based on the vessel navigation data of the target channel section and the dynamic water depth field of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel, determine the status label of each predicted time point in the initial time series of the initial dynamic water depth data of each target channel. The status labels include safety, caution, and danger. Write-back time series unit: Based on the status labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, determine the write-back time series of each target channel.

5. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 4, characterized in that, Write back to the time series unit, including: Initial state sequence sub-unit: Based on the state labels of all predicted time points in the initial time series of the initial dynamic water depth data of each target channel, the initial state sequence of each target channel is determined. Judgment sub-unit: If two adjacent state labels are the same in the initial state sequence of the target river channel, determine the predicted time point corresponding to the first state label in the initial time sequence as the adjustment time point, and discard the predicted time point corresponding to the second state label in the initial time sequence. Adjustment time point subunit: If the status labels of every two adjacent predicted time points in the initial time series of the initial dynamic water depth data of the target river channel are different, the two adjacent predicted time points are determined as adjustment time points respectively; Interpolation time point subunit: Based on the state labels of two adjacent prediction time points with different state labels in the initial time series of the initial dynamic water depth data of the target river channel and the dynamic water depth field, the interpolation time point of the two adjacent prediction time points is determined; Iterative sub-unit: Iterate over all adjustment time points and all interpolation time points for each target river channel until two adjacent adjustment time points or interpolation time points satisfy the first rule or the second rule. The first rule is that the state labels are the same, and the second rule is that when the state labels are different, the time interval between two adjacent adjustment time points or interpolation time points is less than or equal to the preset minimum time interval. Write-back time series sub-unit: Based on all adjustment time points and all interpolation time points of the last iteration for each target channel, determine the write-back time series for each target channel.

6. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 1, characterized in that, Write-back module, including: Electronic navigation data unit: Acquires electronic navigation data based on standard data format from electronic navigation charts. The electronic navigation data includes at least depth point storage data and depth derived data, and acquires depth value encoding rules. The depth derived data includes multiple original derived elements. Predicted water depth value unit: Based on the coordinates of each stored water depth point in the water depth point storage data, spatial interpolation is performed on the dynamic water depth field at each adjustment time point or interpolation time point in the write-back time series to determine the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series. Encoded water depth value unit: Based on the water depth value encoding rules, the predicted water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series is encoded to determine the encoded water depth value of each stored water depth point at each adjustment time point or interpolation time point in the write-back time series; Depth Point Replacement Data Unit: Locates the record identifier of each stored depth point in the depth point storage data, writes the coded depth value of each stored depth point at each adjustment time point or interpolation time point in the write-back time series back to the record position of the corresponding stored depth point, and determines the depth point replacement data based on the standard data format for each adjustment time point or interpolation time point in the write-back time series. Derivative Replacement Data Unit: Based on the water depth point replacement data of each adjustment time point or interpolation time point in the write-back time series, generate derivative replacement data of each adjustment time point or interpolation time point in the write-back time series based on the standard data format. The derivative replacement data includes multiple replacement derivative elements. Update channel data unit: Update electronic channel data based on standard data format depth point replacement data and derived replacement data for each adjustment or interpolation time point in the write-back time series, and determine the updated channel data for each adjustment or interpolation time point in the write-back time series.

7. The electronic navigation chart navigation element rendering system based on graphics acceleration according to claim 1, characterized in that, The rendering module includes: Rendering Method Unit: Obtains a set of derived feature types and a graphics-accelerated rendering method for each feature type in the set of derived feature types. The set of derived feature types includes at least point cloud features, point symbol features, line features, and complex building features. Updated Channel Chart Unit: A rendering method based on the feature type of each replacement derivative element in the derivative replacement data of the updated channel data at each adjustment time point or interpolation time point in the write-back time series is used to render the replacement derivative elements and determine the updated channel chart at each adjustment time point or interpolation time point in the write-back time series.

8. A graphics-accelerated electronic navigation chart navigation element rendering platform, characterized in that, Used to execute the graphics-accelerated electronic navigation chart navigation element rendering system as described in any one of claims 1 to 7.

Images