Intelligent navigation data processing system for combined fleet

Abstract

The invention relates to the technical field of combined fleet navigation, and discloses a combined fleet intelligent navigation data processing system, which comprises a data acquisition unit used for acquiring navigation data and ship lock data of each ship of a combined fleet in real time, fusing the preprocessed navigation data and ship lock data to obtain a comprehensive data set, through dynamic data fusion and self-adaptive grid division, the spatial-temporal characteristics of a flow velocity abrupt change area during opening and closing of a ship lock are captured in real time, the limitation that a traditional algorithm depends on a static environment is broken through, a dynamic power distribution strategy is generated based on flow field characteristics and navigation data, and precise power matching is achieved in combination with a propeller efficiency priority and a self-adaptive weight algorithm. The response time is shortened, the cooperation efficiency is improved, course deviation is avoided, a multi-stage early warning and time difference judgment mechanism is constructed, the power margin is evaluated in real time, the weight, the navigational speed and the formation are adjusted according to the risk level, it is guaranteed that power distribution adjustment of the whole fleet does not lag behind water flow changes, and normal course of the fleet is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of combined fleet navigation, and particularly relates to an intelligent navigation data processing system for a combined fleet. Background Art

[0002] Currently, in inland waterway shipping, in order to meet the shipping requirements, combined fleets are usually used for shipping. When passing through a lock, multiple ships form a fleet and wait to pass through the lock. Before entering the narrow lock chamber, due to the narrow lock and complex water flow, each ship dynamically adjusts its power according to its own load and resistance to ensure that the combined fleet can pass through the lock gate smoothly.

[0003] When the existing combined fleet passes through a lock, the multi-objective particle swarm optimization algorithm is used for power distribution, and the coordinated control of multiple thrusters is realized by integrating data such as ship load, thruster efficiency, and energy consumption. However, this algorithm depends on a static environment, such as a preset water flow velocity and a constant direction. The sudden turbulent flow generated during the opening and closing of the lock has strong nonlinear and unsteady characteristics, resulting in a flow velocity mutation area. The fleet often lags behind the water flow change when adjusting the power distribution after the data processing is completed, and ultimately may cause the fleet's course deviation. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an intelligent navigation data processing system for a combined fleet, which solves the above problems.

[0005] The above technical object of the present invention is achieved through the following technical solutions:

[0006] An intelligent navigation data processing system for a combined fleet includes:

[0007] A data acquisition unit, configured to collect the navigation data and lock data of each ship in the combined fleet in real time, fuse the preprocessed navigation data and lock data, and obtain a comprehensive data set;

[0008] A positioning unit, configured to perform feature extraction and analysis on the comprehensive data set, locate the position and range of the flow velocity mutation area when the lock is opened and closed, and generate a flow field feature parameter set;

[0009] An allocation unit, configured to analyze the flow field feature parameter set and the navigation data, and generate a power distribution instruction;

[0010] A determination unit, configured to calculate the power and flow field time difference for the comprehensive data set, and determine whether to execute the power distribution instruction based on the power and flow field time difference.

[0011] Further, fusing the preprocessed navigation data and lock data to obtain a comprehensive data set includes:

[0012] Analyze the preprocessed navigation data and lock data to obtain a correlation matrix;

[0013] Based on the correlation matrix, fuse the preprocessed navigation data and lock data to obtain a fused dataset;

[0014] Perform anomaly processing on the fused dataset to obtain a comprehensive dataset.

[0015] Furthermore, perform feature extraction and analysis on the comprehensive dataset to locate the position and range of the flow velocity mutation area during lock opening and closing, and generate a set of flow field characteristic parameters, including:

[0016] Separate the spatio-temporal fluctuation modes of the flow field from the comprehensive dataset to generate a dynamic flow field characteristic vector;

[0017] Extract the flow velocity information of each spatial position and time point in the dynamic flow field characteristic vector, and use the central difference method to calculate the modulus of the flow velocity gradient at each spatial point in adjacent spatial and time steps, and set three gradient thresholds: high, medium, and low;

[0018] When the modulus of the flow velocity gradient > high gradient threshold, divide the flow field space into a high-speed change area. For the high-speed change area, reduce the grid spacing to half of the original spacing;

[0019] When the low gradient threshold < modulus of the flow velocity gradient < medium gradient threshold, divide the flow field space into a medium-speed change area, and adjust the grid spacing to three-quarters of the original spacing;

[0020] When the modulus of the flow velocity gradient < low gradient threshold, divide the flow field space into a low-speed change area, and increase the grid spacing to 2 times the original spacing for sparse processing;

[0021] Use the Delaunay triangulation algorithm to regenerate grid nodes according to the new grid division, determine the three-dimensional spatial coordinates of the grid nodes, analyze the connection relationship between nodes, construct a grid correlation matrix, record adjacent grid cell information, and finally obtain a dynamic grid topology;

[0022] Decompose the dynamic grid topology into a steady-state flow velocity field and a transient turbulent flow field, and generate a steady-state flow velocity distribution matrix and a transient vortex intensity matrix respectively;

[0023] Compare the vortex flow velocity of the transient vortex intensity matrix with the reference value of the steady-state flow velocity distribution matrix to generate a flow velocity mutation difference field, specifically as follows: Normalize the vortex flow velocity and the reference flow at each grid node, then calculate the sum of the squares of the differences, and finally take the square root to obtain the overall measure of the normalized flow velocity difference at all nodes. Divide the overall measure by the number of grid divisions to average the overall difference degree to each grid and obtain the spatially averaged flow velocity difference parameter;

[0024] Normalize the time change rate of the transient vortex flow velocity to obtain the ratio of the current flow velocity change rate to the maximum change rate. Then, normalize the ship length and the effective length of the lock respectively and divide them to obtain the ratio of the ship length to the effective length of the lock. Multiply the ratio of the current flow velocity change rate to the maximum change rate by the ratio of the ship length to the effective length of the lock and add 1 to form a correction factor;

[0025] Multiply the spatially averaged flow velocity difference parameter by the correction factor to obtain the flow velocity mutation difference field.

[0026] Furthermore, perform feature extraction and analysis on the comprehensive data set to locate the position and range of the flow velocity mutation area during lock opening and closing, and generate a flow field feature parameter set, which also includes:

[0027] Identify the boundary of the mutation area of the flow velocity mutation difference field to obtain the flow velocity mutation boundary coordinate set;

[0028] Calculate the flow velocity gradient components of each grid node in the steady-state flow velocity distribution matrix in the three spatial directions of transverse, longitudinal, and vertical through the central difference method. Combine these three components into a spatial gradient vector and calculate the magnitude of this vector to obtain the spatial gradient magnitude;

[0029] Extract the transient flow velocity data of each node at different times from the dynamic grid topology structure, and calculate the decay rate and decay multiple based on the transient flow velocity data;

[0030] Combine the spatial gradient magnitude, decay rate, and decay multiple to generate a turbulent intensity gradient parameter set;

[0031] Fuse the flow velocity mutation boundary coordinate set and the turbulent intensity gradient parameter set to generate a flow field feature parameter set.

[0032] Furthermore, analyze the flow field feature parameter set and the navigation data to generate a power distribution instruction, including:

[0033] Analyze the convective flow field characteristic parameter set to obtain the peak velocity difference in the velocity mutation area and the turbulence influence gradient, as follows: In the flow field, extract the maximum value of the velocity gradients in three spatial directions, multiply it by the draft of the ship, divide the product of the maximum velocity gradient and the draft by the average steady-state velocity to obtain the relative parameter of velocity gradient - draft;

[0034] Divide the draft of the ship by the effective water depth of the lock to obtain the water depth ratio, divide the lock opening and closing time by the fleet passing-through lock processing time to obtain the time ratio, multiply the two ratios and then add 1 to obtain the lock - time correlation correction coefficient;

[0035] Divide the difference between the peak velocity in the velocity mutation area and the current ship speed by the maximum steady-state velocity and then add 1 to obtain the velocity difference - steady-state velocity correction coefficient;

[0036] Multiply the relative parameter of velocity gradient - draft, the lock - time correlation correction coefficient, and the velocity difference - steady-state velocity correction coefficient to obtain the turbulence influence gradient;

[0037] Subtract the current ship speed from the peak velocity in the velocity mutation area to obtain the peak velocity difference;

[0038] Calculate the peak velocity difference and the ship draft data to obtain the power increment;

[0039] Combine the turbulence influence gradient and the power increment and perform correction to obtain the corrected power increment;

[0040] Calculate the current output power of the main engine and the corrected power increment to obtain the power margin coefficient. If the power margin coefficient < 1, trigger the hierarchical warning mechanism, as follows: Correct the basic power demand of the current working condition of each ship in the fleet, then sum to obtain the total corrected power demand of the fleet, subtract the total power demand from the current total output power of the main engine to obtain the difference between the current output power of the main engine and the total corrected power demand of the fleet, and then divide the difference between the current output power of the main engine and the total corrected power demand of the fleet by the rated reserve power of the main engine to obtain the power margin coefficient.

[0041] Furthermore, analyze the convective flow field characteristic parameter set and the navigation data to generate a power distribution instruction, which also includes:

[0042] Extract the navigation data to generate a power distribution feasibility matrix;

[0043] Fuse the thrust efficiency coefficient and the power demand increment of the power distribution feasibility matrix according to the weight, use the weighted summation algorithm to obtain the comprehensive efficiency value of each thruster, and in the complex flow field, arrange the thrusters from high to low according to the comprehensive efficiency value through the quick sorting algorithm to finally obtain the thruster efficiency priority;

[0044] Fuse the corrected power increment, power margin coefficient, and power distribution feasibility matrix to generate a power distribution decision parameter set that includes the comprehensive performance indicators of each thruster;

[0045] According to the thruster efficiency priority and the power distribution decision parameter set, use the adaptive weight algorithm to dynamically allocate the thruster power adjustment amount for each ship. The initial weights are allocated according to the thruster efficiency priority of the ship, and then the weights are quantitatively adjusted for different ship thrusters based on the flow field type and flow velocity grading. According to the adjusted weights, the total power adjustment amount is proportionally allocated to the thrusters of each ship to generate a power distribution instruction.

[0046] Furthermore, if the power margin coefficient 1, trigger a hierarchical warning mechanism, including:

[0047] When 0.8 power margin coefficient 1, it is determined that there is a critical risk of remaining power of the main engine in the combined fleet, trigger a first-level warning, and the yellow lights of the combined fleet flash;

[0048] When 0.5 power margin coefficient 0.8, it is determined that there is a risk of shortage of remaining power of the main engine in the combined fleet, trigger a second-level warning, and the orange lights of the combined fleet flash;

[0049] When the power margin coefficient 0.5, it is determined that there is a serious shortage risk of remaining power of the main engine in the combined fleet, trigger a third-level warning, and the red lights of the combined fleet flash.

[0050] Furthermore, calculate the power and flow field time difference from the comprehensive data set, and determine whether to execute the power distribution instruction based on the power and flow field time difference, including:

[0051] Extract the flow field time parameter and power time parameter of the flow field change from the comprehensive data set;

[0052] Calculate the difference between the flow field time parameter and the power time parameter to obtain the power and flow field time difference;

[0053] If the difference between the power and the flow field time ≤ 0, it means that the effective time of the power adjustment lags behind the influence time of the flow field change, determine not to execute the power distribution instruction, and execute the forced trigger instruction;

[0054] If the power and flow field time difference > 0, it means that the power adjustment can take effect before the flow field mutation affects the ship, and determine to execute the power distribution instruction.

[0055] Furthermore, the hierarchical warning mechanism also includes:

[0056] At the first-level warning: Based on the current flow field type, the current flow field type is divided into cross flow, counter current and downstream flow. On the basis of the original weight adjustment rule, the weight of the high-efficiency thrusters of the ships in the top 50% of the priority is additionally increased by 0.1, and at the same time, the weight of the low-efficiency thrusters of the ships in the bottom 50% of the priority is reduced equally;

[0057] Temporarily reduce the target speed by 5% - 10%;

[0058] At the second-level warning: Temporarily increase the weight of the thrust efficiency coefficient to 0.7 and reduce the weight of the power demand increment to 0.3;

[0059] For the ship thrusters in the leeward side or auxiliary position, the weight upper limit is reduced to 0.2, and the weight of the thrusters on the windward side or leading ship is increased to 0.7 - 0.8, specifically as follows:

[0060] Counter current or cross flow scenario: Adjust the weight of the thrusters of the main ship on the windward side to the upper limit of the rule, and the weight of the thrusters of the auxiliary ship on the leeward side is reduced to a minimum of 0.05;

[0061] Downstream flow scenario: The weight of the thrusters of the front-row ships is reduced below 0.3, and the weight of the thrusters of the escort ships on both sides is increased to 0.35. Control the course through side thrusters to reduce the consumption of main thrust power;

[0062] The formation changes to a "single column" formation, and the leading ship's course is slightly adjusted by 5° - 10° to avoid the strong current direction and reduce the overall water flow resistance;

[0063] Allow the speed to be temporarily 10% - 20% lower than the target speed, and give priority to ensuring that the power system is not overloaded;

[0064] At the third-level warning: Only activate 1 ship thruster with the highest efficiency priority, directly set its weight to 1, and set the remaining thrusters to 0 to satisfy the weight conservation;

[0065] Immediately perform a stop operation, turn off all non-essential loads, and only maintain the operation of navigation and communication equipment;

[0066] Send a distress signal to the nearby port or maritime department;

[0067] Among them, for the third-level warning the second-level warning the first-level warning, the high-level warning measures automatically cover the low-level warning measures.

[0068] Furthermore, execute the forced trigger instruction, including:

[0069] Send an alarm to remind the operator to manually force an early trigger of power adjustment and enforce the power distribution instruction.

[0070] In summary, the present invention mainly has the following beneficial effects:

[0071] Through dynamic data fusion and adaptive grid division technology, a flow field analysis system for real-time response to sudden turbulence is constructed. First, through the correlation matrix, the navigation data and the ship lock data are deeply fused. After removing the interference of abnormal data, a comprehensive data set is formed to ensure the integrity and reliability of the input information. Then, the central difference method is used to calculate the velocity gradient in real time. According to high, medium, and low thresholds, the grid density is dynamically adjusted. In the area of sudden velocity change, the grid spacing is refined to one-half of the original spacing, and the low-velocity change area is sparsely processed synchronously. Through Delaunay triangulation, a dynamic grid topology structure is constructed, effectively separating the steady-state velocity field and the transient turbulence field, generating a set of flow field characteristic parameters including the coordinates of the sudden change boundary and the turbulence attenuation characteristics, accurately positioning the temporal and spatial range of the sudden velocity change during the opening and closing of the ship lock. This mechanism breaks through the static environment assumption, can capture sudden turbulence with a large rate of change in velocity gradient in real time, provides high-precision dynamic flow field parameters for power distribution, and fundamentally solves the problem of the lag in the adaptation of traditional algorithms to unsteady flow fields.

[0072] By analyzing the peak velocity difference and the turbulence influence gradient in the area of sudden velocity change, combined with the ship's draft data, the basic power increment is calculated. Then, through the turbulence influence gradient, the weight is corrected to make the power adjustment amount accurately match the intensity of the flow field disturbance. Using the power distribution feasibility matrix, key parameters such as the propeller efficiency coefficient and the power demand increment are integrated, and the propeller efficiency priority is generated through the weighted summation algorithm. In the countercurrent scenario, the weight of the main propulsion ship's propeller can be increased, while suppressing the power consumption of the inefficient propellers of the auxiliary ships on the backflow side. Based on the adaptive weight algorithm, the power adjustment amount is dynamically distributed. The initial weight is set according to the efficiency priority, and then it is adjusted secondly according to the flow field type (cross flow, downstream flow, countercurrent flow). For example, in the downstream flow scenario, the weight of the propellers of the front-row ships is reduced, and that of the escort ships on both sides is increased, realizing the optimal ratio of propulsion power, shortening the overall power distribution response time of the fleet, and improving the cooperative efficiency of the propellers, effectively avoiding the risk of the fleet's course deviation caused by the lag in power distribution.

[0073] Through the precise calculation of the time difference between power and flow field and the hierarchical early warning strategy, a full-process guarantee system covering risk early warning, dynamic adjustment, and emergency handling is established. By extracting the time parameters of the flow field change and the power response time parameters in real time, when the time difference ≤ 0, it is determined that the power adjustment lags, and the forced manual intervention mechanism is immediately triggered. The operator is prompted to intervene in advance through audible and visual alarms to ensure that the power configuration is completed before the flow field mutation affects the ship. For the problem of insufficient power margin, a three-level gradient early warning mechanism is designed: at the first-level early warning, the propeller weight is dynamically adjusted and the target ship speed is moderately reduced to reserve a part of the power margin. At the second-level early warning, the weight distribution is re-performed, and at the same time, the fleet formation is changed to a "single column" with better flow resistance performance, and the course is adjusted to reduce the water flow resistance. At the third-level early warning, the single-point control of the most efficient propeller is enabled, non-essential loads are synchronously turned off, and a distress signal is sent, forming a closed-loop control from risk early warning to emergency handling, overall reducing the overload risk of the power system and controlling the course deviation of the fleet under sudden strong turbulent flow conditions, significantly improving the safety and reliability of the combined fleet passing through the lock process. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 is a block diagram of an intelligent navigation data processing system for a combined fleet according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0075] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0076] Reference Figure 1 , an intelligent navigation data processing system for a combined fleet, comprising:

[0077] A data acquisition unit for real-time collecting the navigation data and lock data of each ship in the combined fleet, fusing the preprocessed navigation data and lock data to obtain a comprehensive data set;

[0078] A positioning unit for feature extraction and analysis of the comprehensive data set, positioning the position and range of the flow velocity mutation area during lock opening and closing, and generating a flow field feature parameter set;

[0079] An allocation unit for analyzing the flow field feature parameter set and the navigation data to generate a power allocation instruction;

[0080] A determination unit for calculating the comprehensive data set to obtain the time difference between power and flow field, and determining whether to execute the power allocation instruction based on the time difference between power and flow field.

[0081] Real-time data integration is achieved through the data acquisition unit, overcoming the problem of lagging data integration in the prior art and providing an accurate and comprehensive data basis for subsequent processing. The flow velocity mutation area is accurately identified through the positioning unit, effectively coping with the complex flow field changes during the opening and closing of the ship lock and avoiding the deviation of the fleet's heading caused by sudden turbulence. The distribution unit generates an adaptive power distribution instruction based on the flow field characteristics and navigation data, and the determination unit ensures the timeliness of power adjustment through time difference verification, significantly improving the navigation stability and safety of the fleet in the dynamic flow field. Moreover, this system breaks through the limitations of traditional static algorithms and constructs a closed-loop control system of precise perception-intelligent decision-making-dynamic regulation, greatly enhancing the reliability and efficiency of the lockage operation of the combined fleet and enabling the collaborative operation of ships in complex water flow environments.

[0082] In one case of this embodiment, the preprocessed navigation data and ship lock data are fused to obtain a comprehensive data set, including:

[0083] The navigation data includes: ship position data, ship speed data, heading data, ship attitude data, ship draft data, main engine power data, and propeller speed data;

[0084] The ship lock data includes: ship lock water level data, ship lock flow velocity data, gate opening and closing time data, ship lock flow field characteristic time data, ship lock structure parameter data, and ship lock water flow sediment concentration data;

[0085] The preprocessed navigation data and ship lock data are analyzed to obtain a correlation matrix, specifically including: for the preprocessed navigation data and ship lock data, the samples are aligned according to the time stamp to construct a unified data set. The Pearson correlation coefficient algorithm is used to calculate the correlation coefficients between the variables of the navigation data (ship position, ship speed) and the variables of the ship lock data (ship lock water level, flow velocity) respectively. These coefficients are arranged in rows and columns corresponding to the variables, and the diagonal elements are set to 1 to form a correlation matrix;

[0086] Based on the correlation matrix, the preprocessed navigation data and ship lock data are fused to obtain a fused data set, specifically including: the fusion work is carried out according to the Pearson correlation coefficients in the correlation matrix. For the correlation coefficients between the variables of the navigation data and the variables of the ship lock data in the correlation matrix, the threshold is set to 0.7. If the absolute value of the correlation coefficient is greater than or equal to this threshold, it means that the two variables are strongly correlated. At this time, the weighted average method is used to fuse the data, and the correlation coefficient is used as the weight to calculate the weighted average value as the fused data value. If the absolute value of the correlation coefficient is less than the threshold, it indicates that the correlation is weak, and the original data of these two variables are retained respectively without fusion processing. The data of all variables are integrated in chronological order, and the processed and retained data are arranged in an orderly manner to form a fused data set that includes both the characteristics of the navigation data and the characteristics of the ship lock data;

[0087] Perform anomaly processing on the fused dataset to obtain a comprehensive dataset, which specifically includes: for the fused dataset, use the box plot method for processing. First, calculate the first quartile (lower quartile) and the third quartile (upper quartile) of each variable's data to determine the distribution interval of the middle 50% of the data. Then, calculate the reasonable range boundaries of the data fluctuation through this distribution interval, and determine the data outside the lower boundary or above the upper boundary as outliers. For outliers, if the sample size of the dataset is less than 30, directly eliminate it. If the sample size is greater than or equal to 30, use the median of the same variable's data to replace and correct the outliers. Finally, obtain the cleaned comprehensive dataset.

[0088] By detailed partitioning and classification of various navigation data such as ship position and speed and lock data such as lock water level and flow rate, the comprehensiveness and accuracy of the original data are ensured. After preprocessing, use the Pearson correlation coefficient algorithm to construct a correlation matrix to quantify the correlation of each variable between navigation data and lock data, providing an objective basis for subsequent fusion. Based on the correlation matrix, set reasonable thresholds and use the weighted average method to fuse strongly correlated variables, fully considering the importance and correlation of each variable, so that the fused dataset not only retains the characteristics of navigation data but also covers the characteristics of lock data, effectively integrating multi-source information, providing high-quality and comprehensive data support for subsequent operations such as power distribution, and greatly enhancing the intelligent decision-making basis for the combined fleet to pass through the lock, improving the safety and stability of the fleet passing through the lock.

[0089] By performing anomaly processing on the fused dataset using the box plot method, determine the reasonable range boundaries based on the data quartiles, and accurately identify outliers. For different sample size anomaly situations, adopt strategies of elimination or replacement and correction with the median respectively, fully considering the influence of the data sample size on the anomaly processing effect, which can not only effectively remove noise data but also avoid data information loss caused by too small sample size. The finally obtained comprehensive dataset has high accuracy and strong reliability, which provides a solid guarantee for subsequent steps such as feature extraction, analysis of the comprehensive dataset, and generation and determination of power distribution instructions, ensuring that the combined fleet can make intelligent navigation decisions based on accurate and reliable data in a complex lock environment, effectively reducing the navigation risks caused by data anomalies, and improving the safety and stability of the fleet passing through the lock.

[0090] In one case of this embodiment, perform feature extraction and analysis on the comprehensive dataset to locate the position and range of the flow velocity mutation area when the lock opens and closes, and generate a flow field feature parameter set, including:

[0091] Separate the spatio-temporal fluctuation modes of the flow field from the comprehensive dataset to generate dynamic flow field feature vectors, specifically including: extracting data related to the flow field (lock velocity, water level) from the comprehensive dataset, constructing a spatio-temporal matrix with spatial positions as rows and time points as columns, performing singular value decomposition on the spatio-temporal matrix to obtain three matrices: the left singular matrix represents the spatial mode, the right singular matrix represents the time mode, and the singular value matrix represents the energy proportion of each mode. Sort the singular values, select the top 10 modes with high energy proportion as the main spatio-temporal fluctuation modes, and extract the vectors of these modes from the corresponding matrices to form the dynamic flow field feature vectors;

[0092] Extract the velocity information of each spatial position and time point in the dynamic flow field feature vector, and use the central difference method to calculate the modulus of the velocity gradient at each spatial point in adjacent space and time steps, and set three gradient thresholds: high, medium, and low;

[0093] When the modulus of the velocity gradient is greater than the high gradient threshold, divide the flow field space into a high-speed change area, and for the high-speed change area, reduce the grid spacing to half of the original spacing;

[0094] When the low gradient threshold is less than the modulus of the velocity gradient and less than the medium gradient threshold, divide the flow field space into a medium-speed change area, and adjust the grid spacing to three-quarters of the original spacing;

[0095] When the modulus of the velocity gradient is less than the low gradient threshold, divide the flow field space into a low-speed change area, and increase the grid spacing to 2 times the original spacing for sparse processing;

[0096] Based on the initial grid division, clarify the three-dimensional spatial coordinates of the existing grid nodes, regenerate the grid nodes according to the new grid division through the Delaunay triangulation algorithm, determine the coordinate positions of these new nodes in the three-dimensional space. On this basis, analyze the connection relationship between the old and new nodes, clarify the adjacent situation of each node, and construct a grid association matrix according to the clear node connection relationship. This matrix accurately records the connection state of each node with other nodes, marked as 1 if adjacent, otherwise 0, and comprehensively records the information of adjacent grid cells, covering the shared faces, edges, and nodes between cells. Finally, integrate all node coordinates, connection relationships, and cell adjacent information to construct a dynamic grid topology structure that can dynamically reflect the grid changes;

[0097] Decompose the dynamic grid topology structure into a steady-state flow velocity field and a transient turbulent flow field, and generate a steady-state flow velocity distribution matrix and a transient vorticity intensity matrix respectively, specifically including: obtaining the velocity data of each node at different times from the dynamic grid topology structure;

[0098] For the steady-state flow velocity field, the average flow velocity at each grid node over the entire time period is calculated by the time-averaging method. These average flow velocities are used as matrix elements to construct a steady-state flow velocity distribution matrix, where each element represents the reference flow velocity value of the corresponding grid node;

[0099] For the transient turbulent flow field, the transient component is obtained by subtracting the corresponding reference flow velocity from the flow velocity at each node at each moment. The vorticity tensor is calculated based on the transient component. The Q-criterion is used to analyze the vorticity tensor to identify the vortex core region and obtain the core position coordinates of the vortex core region. For each identified vortex core region, the vorticity magnitude is calculated according to the vorticity tensor, and this vorticity magnitude is used as the rotation intensity of this vortex. The DBSCAN clustering algorithm is used to cluster the vortex core regions, and the vortex cores that are close to each other in space are grouped into one class to obtain the clustering result. Combining time-series correlation, the clustering results at different moments are analyzed to track the entire process of each vortex from generation to extinction, determine its life cycle, and organize the core position coordinates, rotation intensity, and life cycle of the vortex into a matrix form to generate a transient vortex intensity matrix;

[0100] The vortex flow velocity in the transient vortex intensity matrix is compared with the reference value in the steady-state flow velocity distribution matrix to generate a flow velocity mutation difference field. The specific calculation formula is as follows: ;

[0101] In the formula, represents the flow velocity mutation difference field, and respectively represent the number of divisions of the grid in the transverse and longitudinal directions, represents the vortex flow velocity at the grid node in the th row and th column in the transient vortex intensity matrix, represents the maximum value of the vortex flow velocity, represents the reference flow velocity of the corresponding grid node in the steady-state flow velocity distribution matrix, represents the maximum value of the reference flow velocity, represents the time change rate of the transient vortex flow velocity, which is used to reflect the speed of the flow velocity change, represents the maximum value of the time change rate, represents the ship length, represents the effective length of the lock, represents the length reference quantity, is and have the same dimension;

[0102] Through the decomposition based on the spatio-temporal matrix, it is possible to separate the main fluctuation modes with high energy proportion from the complex lock flow velocity and water level data, effectively filter out noise and retain key dynamic features. Combining with the central difference method to calculate the modulus of the flow velocity gradient, the grid spacing is dynamically adjusted according to the gradient threshold, so that the grid density in the high-speed change area (such as the strong shear flow area during lock opening and closing) is automatically encrypted, while the low-speed change area is sparsely processed. While ensuring the calculation accuracy, the calculation efficiency is significantly improved. Through the division of the adaptive grid, it not only avoids the calculation redundancy caused by the global dense grid, but also prevents the loss of key details caused by the sparse grid, and is especially suitable for the strong non-uniform flow scenario caused by the sudden change of boundary conditions in the flow field near the lock. In addition, the dynamic grid topology structure constructed based on Delaunay triangulation can accurately record the node connection relationship and element adjacent information, providing a solid data basis for the analysis of the flow field and ensuring the accuracy and reliability of the positioning of the flow velocity mutation area.

[0103] By constructing a steady-state flow velocity distribution matrix through the time-averaging method, the reference flow state of the flow field can be clearly presented, providing a stable reference framework for identifying abnormal flows. Through the analysis of transient vortices, the accurate identification, intensity quantification and life cycle tracking of the core area of the vortices are realized, effectively capturing the complex vortex motions (such as shear vortices and recirculation area vortices near the gate) caused by the sudden change of flow velocity during the lock opening and closing process. And the flow velocity mutation difference field generated by comparing the vortex flow velocity with the steady-state reference value brings in parameters such as the ship length and the effective length of the lock, making the flow field analysis closely combined with the actual navigation requirements. This difference field can not only locate the spatial position of the flow velocity mutation (such as the high-gradient area upstream and downstream of the gate), but also provide an analysis basis for the safety assessment during ship lock passage and the optimization of lock scheduling strategies by quantifying the coupling effect of the time rate of flow velocity change and the characteristic length, facilitating the dynamic adjustment of lock operation parameters.

[0104] In one case of this embodiment, feature extraction and analysis are performed on the comprehensive data set to locate the position and range of the flow velocity mutation area during lock opening and closing, and a set of flow field characteristic parameters is generated, and it also includes:

[0105] Identify the boundaries of the mutation regions in the flow velocity mutation difference field to obtain the coordinate set of the flow velocity mutation boundaries, which specifically includes: for the flow velocity mutation difference field, first perform edge detection using the Sobel operator. By calculating the flow velocity change rates of each grid point in the horizontal and vertical directions, obtain the gradient magnitude (reflecting the mutation magnitude) and gradient direction (reflecting the mutation direction) of each point. Then, perform adaptive threshold segmentation on the gradient magnitude matrix using the Otsu algorithm. By analyzing the gray distribution of the data histogram, set the threshold, and mark the regions with gradient values higher than the threshold as candidate regions for flow velocity mutation, excluding the stable regions with low gradients. For the marked candidate regions for flow velocity mutation, use the Marching Squares algorithm to extract the boundary contours, and scan the grid cells row by row and column by column to detect the transition points where the gradient values on the cell edges change from higher than the threshold to lower than the threshold. Through linear interpolation, based on the coordinates of the two endpoints of the edge, calculate the exact coordinates of the boundary points. Integrate the gradient magnitude and gradient direction of this point calculated by the Sobel operator before with the newly calculated coordinates and record them in order, thus forming a coordinate set of flow velocity mutation boundaries that includes coordinates, mutation magnitude, and direction;

[0106] Calculate the flow velocity gradient components of each grid node in the steady-state flow velocity distribution matrix in three spatial directions: horizontal, vertical, and vertical. Combine these three components into a spatial gradient vector. Then, square the flow velocity gradient components in the three directions respectively and add them together, and take the arithmetic square root of the sum. The scalar value obtained is the spatial gradient amplitude;

[0107] Extract the transient flow velocity data of each node at different times from the dynamic grid topology. Obtain the pulsating velocity by subtracting the reference flow velocity (steady-state average flow velocity) of the corresponding node from the transient flow velocity. Take the root mean square value of the pulsating velocity at each time to generate a turbulent intensity sequence that changes with time for each node. For the turbulent intensity sequence, use the sliding window difference method to calculate the decay rate of adjacent time steps (i.e., the difference between the turbulent intensity at the next time step and the previous time step divided by the time interval), and fit the overall trend of the sequence using the least squares method. Extract the slope of the fitted straight line as the overall decay rate of the turbulent intensity of this node. At the same time, calculate the ratio of the initial value to the termination value of the sequence as the decay multiple to quantify the decay amplitude of the turbulent intensity over time;

[0108] In the integration stage, use the spatial coordinates ( ) of each grid node as the positioning reference, take the spatial gradient amplitude (reflecting the spatial variation characteristics of the flow velocity), decay rate (reflecting the time decay speed), and decay multiple (reflecting the time decay amplitude) of this node as characteristic attributes, and organize them in the matrix form of "one node per row and one parameter per column". Finally, generate a turbulent intensity gradient parameter set that includes node position information, spatial gradient characteristics of the flow velocity, and turbulent time decay characteristics;

[0109] Fuse the flow velocity mutation boundary coordinate set and the turbulent intensity gradient parameter set to generate a flow field characteristic parameter set, which specifically includes: based on the boundary point coordinates of the flow velocity mutation boundary coordinate set, extract the coordinates of the boundary points and construct an array. According to the spatial distribution of the boundary point coordinates, interpolate and calculate the spatial gradient amplitude, attenuation rate, and attenuation multiple of each grid node in the turbulent intensity gradient parameter set, so as to obtain the turbulent intensity parameter values at the corresponding positions of the boundary points. Integrate the spatial coordinates, mutation magnitude, and mutation direction of the boundary points with the interpolated turbulent intensity parameters. Taking the boundary points as the spatial positioning reference, arrange the mutation magnitude, mutation direction, spatial gradient amplitude, attenuation rate, and attenuation multiple of this point in sequence to generate a flow field characteristic parameter set containing node position information, flow velocity spatial gradient characteristics, and turbulent time attenuation characteristics.

[0110] Perform edge detection on the flow velocity mutation difference field through the Sobel operator, accurately calculate the gradient magnitude and direction of each grid point, then use the Otsu algorithm for adaptive threshold segmentation to effectively distinguish the flow velocity mutation and stable regions. Finally, use the Marching Squares algorithm to extract the boundary contour and accurately calculate the boundary point coordinates to form a flow velocity mutation boundary coordinate set containing coordinates, mutation magnitude, and direction. At the same time, use the central difference method to calculate the spatial gradient of the steady flow velocity distribution matrix, and combine the sliding window difference method and the least squares method to calculate the attenuation rate and attenuation multiple of the turbulent intensity, generating a comprehensive turbulent intensity gradient parameter set. This realizes the refined extraction of flow field characteristics, provides a high-precision data basis for the subsequent generation of the flow field characteristic parameter set, improves the perception ability of the combined fleet for potential risk areas in the complex lock environment, helps to formulate more accurate navigation strategies, and ensures the safe passing of the fleet through the lock.

[0111] Through the deep fusion of the flow velocity mutation boundary coordinate set and the turbulent intensity gradient parameter set, generate a comprehensive flow field characteristic parameter set, and based on the boundary point coordinates, interpolate and calculate the turbulent intensity gradient parameters, integrating the spatial coordinates, mutation magnitude, mutation direction, and turbulent intensity parameters of the boundary points to form a complete parameter set containing node position information, flow velocity spatial gradient characteristics, and turbulent time attenuation characteristics. This comprehensively integrates the spatio-temporal characteristics of the flow field and also realizes the organization of multi-dimensional flow field information, providing richer and more accurate flow field characteristic data for ship power distribution and navigation decision-making. Based on this flow field characteristic parameter set, the ship can more accurately predict the flow field changes, adjust the power distribution and navigation attitude in advance, effectively respond to the flow velocity mutation during the opening and closing of the lock, reduce the navigation risk, and improve the passing efficiency and safety of the combined fleet in the complex lock environment.

[0112] In one case of this embodiment, analyze the flow field characteristic parameter set and the navigation data to generate a power distribution instruction, including:

[0113] Analyze the set of convective flow field characteristic parameters to obtain the peak velocity difference and turbulent flow influence gradient in the velocity mutation area. The specific calculation formulas are as follows: ;

[0114] In the formula, represents the turbulent flow influence gradient, which is used to represent the velocity change rate from the edge to the center of the velocity mutation area. , and respectively represent the velocity gradients in three spatial directions. represents the average value of the steady-state velocity. represents the draft of the ship. represents the effective water depth of the lock. represents the opening and closing time of the lock. represents the handling time of the fleet passing through the lock. represents the peak velocity in the velocity mutation area. represents the current speed of the ship. represents the maximum value of the steady-state velocity; ;

[0115] In the formula, represents the peak velocity difference;

[0116] Calculate the peak velocity difference and the ship draft data to obtain the power increment. The specific calculation formula is as follows: ;

[0117] In the formula, represents the power increment. represents the resistance coefficient related to the ship shape and fluid properties. represents that the influence of the draft on the power increment increases in a quadratic relationship;

[0118] Combine the turbulent flow influence gradient and the power increment and perform correction to obtain the corrected power increment. The specific calculation method is as follows: ;

[0119] In the formula, represents the corrected power increment. Among them, the power increment has no dimension. represents the correction weight coefficient;

[0120] Calculate the current output power of the main engine and the corrected power increment to obtain the power margin coefficient. If the power margin coefficient < 1, trigger the hierarchical warning mechanism. The specific calculation formula is as follows: ;

[0121] In the formula, represents the power margin coefficient, which is used to measure the matching degree between the current output power of the main engine and the required power increment. represents the current total output power of the main engine. represents the number of ships in the fleet. represents the basic power demand of the th ship under the current working condition.

[0122] Through the precise analysis of the flow field characteristic parameter set, the peak velocity difference and the turbulent influence gradient in the flow velocity mutation area are obtained. The calculation formula fully considers multi-dimensional parameters such as the ship draft, the effective water depth of the ship lock, the opening and closing time of the ship lock, and the handling time of the fleet passing through the lock, ensuring the accuracy and adaptability of the calculation results. At the same time, the power increment is accurately calculated in combination with the ship draft data, and the correction weight coefficient is introduced to organically combine the turbulent influence gradient with the power increment to obtain the corrected power increment, further improving the scientificity and rationality of the power distribution decision. Finally, by calculating the power margin coefficient and triggering the hierarchical warning mechanism when the power margin coefficient is less than 1, it provides a real-time and accurate power adjustment basis and risk warning for the fleet, effectively ensuring the navigation safety and stability of the fleet in the complex flow field environment.

[0123] Through the generation of power distribution, in this process, the flow field characteristic parameter set and the navigation data are deeply integrated, quantifying the risk degree faced by the ship in the flow velocity mutation area. The calculated peak velocity difference intuitively reflects the difference between the ship speed and the peak velocity in the flow velocity mutation area, providing a key basis for power adjustment. The corrected power increment fully considers the dynamic change of the turbulent influence gradient on the ship's navigation resistance, ensuring that the power distribution not only meets the current navigation requirements but also has a certain safety redundancy. The calculation of the power margin coefficient and the triggering of the hierarchical warning mechanism enable the fleet to predict the risk of insufficient power in advance, take corresponding measures in time, and avoid accidents such as course deviation or collision caused by insufficient power, significantly improving the power management efficiency and navigation safety guarantee level of the combined fleet in the complex ship lock environment.

[0124] In one case of this embodiment, when analyzing the flow field characteristic parameter set and the navigation data to generate a power distribution instruction, it further includes:

[0125] Extract the navigation data to generate a power distribution feasibility matrix (efficiency parameters of different thrusters under the current flow field conditions), specifically including: integrating the current ship speed, target speed, rotational speed of each thruster, water flow speed and flow field characteristic parameter set through a data fusion algorithm, comparing the theoretical thrust and actual thrust of each thruster, calculating the thrust efficiency coefficient, and then combining the draft depth of the ship and the flow velocity distribution in the flow field to calculate the power demand increment of each thruster. Organize the thrust efficiency coefficient and power demand increment of each thruster into a matrix, that is, the power distribution feasibility matrix, where each element represents the efficiency performance of a thruster under the current flow field conditions;

[0126] Fuse the thrust efficiency coefficient and power demand increment of the power distribution feasibility matrix according to the weights, and use the weighted summation algorithm to obtain the comprehensive efficiency value of each thruster. In a complex flow field, the weight of the thrust efficiency coefficient can be set to 0.6, and the weight of the power demand increment can be set to 0.4. And through the quicksort algorithm, arrange the thrusters in descending order according to the comprehensive efficiency value to finally obtain the thruster efficiency priority;

[0127] Fuse the corrected power increment, power margin coefficient and power distribution feasibility matrix to generate a power distribution decision parameter set including the comprehensive performance indicators of each thruster, specifically including: weighted fusion of the thrust efficiency coefficient, power demand increment, corrected power increment and power margin coefficient of each thruster, where the thrust efficiency weight coefficient is 0.3, the power demand increment weight coefficient is 0.2, the corrected power increment weight coefficient is 0.4, and the power margin coefficient weight coefficient is 0.1. Use the weighted summation algorithm to calculate the comprehensive performance indicators of each thruster, and organize the comprehensive performance indicators of all thrusters into a parameter set, that is, the power distribution decision parameter set;

[0128] According to the thruster efficiency priority and the power distribution decision parameter set, use the adaptive weight algorithm to dynamically allocate the power adjustment amount of each ship's thruster. The initial weights are allocated according to the thruster efficiency priority of the ship. Then, based on the flow field type and flow velocity classification, quantitatively adjust the weights of different ship thrusters. According to the adjusted weights, allocate the total power adjustment amount to each ship's thruster in proportion to generate a power distribution instruction;

[0129] The initial weights are allocated according to the thruster efficiency priority of the ship, specifically as follows:

[0130] For a ship formation: The weights decrease in an arithmetic progression to ensure that the sum of all weights is 1. For example:

[0131] For a two-ship formation: The initial weight of the ship thruster with the highest efficiency priority is set to 0.6, and the weight of the ship thruster with the second highest efficiency priority is set to 0.4;

[0132] 3-vessel formation: The weights of the ship propellers with the highest efficiency priority are 0.5, 0.3, and 0.2 in descending order.

[0133] Quantitatively adjust the weights of different ship propellers based on the flow field type and flow velocity classification as follows:

[0134] When in a cross-flow in a complex flow field (perpendicular to the fleet's course):

[0135] At low-speed cross-flow (flow velocity between 1m s - 3m s), the cross-flow has relatively little impact on the ship's course stability. At this time, increase the weight of the propeller of the ship on the upstream side (the ship located in the direction of the cross-flow impact on the fleet) by 0.1, and reduce the weight of the propeller of the ship on the downstream side (the ship in the direction opposite to the cross-flow) by 0.1 (the reduction is evenly distributed);

[0136] Because although there is a cross-flow, the flow velocity is not high, and the ship's own power system can relatively easily overcome the impact of the cross-flow. Slightly adjusting the weight can ensure the ship's course stability and avoid waste of resources;

[0137] For example, in a certain sea area, a 3-vessel freight formation consisting of 1 main cargo ship and 2 auxiliary cargo ships encounters a 2m s lateral water flow during cargo transportation. The main cargo ship is located on the upstream side of the formation, with an initial propeller weight of 0.5. After adjustment, the propeller weight becomes 0.6, while the two auxiliary cargo ships are located on the downstream side, and their propeller weights are reduced from 0.3 and 0.2 to 0.25 and 0.15 respectively. The total weight after adjustment is 0.6 + 0.25 + 0.15 = 1. Through this adjustment, the fleet maintains course stability with minimum power consumption and successfully completes the cargo transportation task;

[0138] At high-speed cross-flow (flow velocity between 3m s - 5m s), the cross-flow will have a significant impact on the ship's course stability. At this time, the weight of the propeller of the ship on the upstream side needs to be increased by 0.3, and the weight of the propeller of the ship on the downstream side is reduced by 0.3 (the reduction is evenly distributed);

[0139] Because the high-velocity cross-flow will generate a significant lateral moment on the fleet, causing the fleet to have a large tendency to deflect. By significantly increasing the weight of the propeller of the ship on the upstream side, the ship on the upstream side can generate enough thrust to overcome the cross-flow resistance and maintain the course of the fleet. At the same time, appropriately reducing the weight of the propeller of the ship on the downstream side can prevent the fleet from deflecting excessively towards the downstream side and ensure the stable navigation of the fleet;

[0140] For example, in a certain sea area, a 3-vessel formation consisting of 1 large freight ship and 2 small escort cargo ships encounters a 4m For the cross-flow of s, the large freight ship is located on the upstream side, with the initial weight of the thruster being 0.6 and increasing to 0.9 after adjustment. Two small escort cargo ships are located on the downstream side, and the weights of their thrusters are reduced from 0.2 and 0.2 to 0.05 and 0.05 respectively. The total weight after adjustment is 0.9 + 0.05 + 0.05 = 1. By significantly enhancing the thrust of the freight ship on the upstream side while reducing the power output of the escort cargo ships on the downstream side, the fleet is successfully prevented from being deflected by the cross-flow.

[0141] When in the countercurrent of a complex flow field (opposite to the fleet's heading):

[0142] When the countercurrent velocity is relatively slow (the velocity is between 1m s - 2m s), the impact on ship navigation is relatively limited. Then, the weight of the thruster of the leading ship in the front row (the first ship in the upstream direction) is increased by 0.1, and the weight of the thruster of the auxiliary ships in the back row is reduced by 0.1 (the reduction is evenly distributed);

[0143] Because the impact of the slow countercurrent on the fleet's navigation is relatively limited, by increasing the weight of the thruster of the leading ship in the front row, its main propulsion power can be enhanced, enabling it to more easily overcome the resistance of the slow countercurrent. And by appropriately reducing the power of the thrusters of the auxiliary ships in the back row, not only can they assist the leading ship in maintaining the formation, but also the energy consumption of the entire fleet can be balanced, achieving efficient utilization of energy;

[0144] For example, a three-ship freight formation consisting of 1 leading cargo ship and 2 auxiliary cargo ships is sailing against the current at 1.5m s in the river. The initial weight of the thruster of the leading cargo ship is 0.5 and it becomes 0.6 after adjustment; the initial weights of the thrusters of the two auxiliary cargo ships are both 0.25 and both are reduced to 0.2 after adjustment. The total weight after adjustment is 0.6 + 0.2 + 0.2 = 1. After adjustment, the leading cargo ship increases its power and successfully propels the fleet forward, while the auxiliary cargo ships reduce their power to reduce energy consumption, and the entire formation completes the cargo transportation with lower fuel consumption;

[0145] When the countercurrent velocity is relatively fast (the velocity is between 2m s - 4m s), it will significantly hinder the ship's forward movement. At this time, the weight of the thruster of the leading ship in the front row is increased by 0.2, and the weights of the thrusters of the auxiliary ships in the back row are each reduced by 0.1;

[0146] Because the strong countercurrent will significantly hinder the fleet's forward movement, the leading ship in the front row needs to provide greater thrust to overcome the stronger countercurrent resistance to ensure that the actual speed of the fleet is close to the target speed. Although the auxiliary ships in the back row can, to a certain extent, assist the fleet in maintaining a stable heading, reducing their power consumption can optimize the overall power distribution and achieve efficient navigation while ensuring the voyage;

[0147] For example, in a certain section of the waterway, a three-ship formation consisting of 1 large-power main cargo ship and 2 medium-sized auxiliary cargo ships encounters a strong countercurrent of 3 m s. The main cargo ship serves as the leading ship, with an initial propeller weight of 0.6, which increases to 0.8 after adjustment. The initial propeller weights of the two auxiliary cargo ships are 0.2 and 0.2 respectively, and the adjusted weights become 0.1 and 0.1. The total adjusted weight is 0.8 + 0.1 + 0.1 = 1. By enhancing the power of the main cargo ship and reasonably reducing the power of the auxiliary cargo ships, the formation successfully maintains its speed, avoids being stranded in dangerous waters due to too slow countercurrent speed, and completes the cargo transportation on time.

[0148] When in the downstream of a complex flow field (in the same direction as the fleet's heading):

[0149] When the downstream flow velocity is relatively fast (the flow velocity is between 3 m s - 5 m s), the ship is prone to excessive speed due to the downstream flow and it is difficult to accurately control the heading. At this time, the propeller weight of the ship in the front row in the downstream direction is reduced by 0.2, and the propeller weights of the two escort ships on both sides are each increased by 0.1 (the total adjustment amount is balanced);

[0150] Because when the downstream flow velocity is relatively fast, the fleet is prone to excessive speed due to the downstream flow, resulting in difficulty in accurately controlling the heading. Reducing the weight of the ship in the front row in the downstream direction can prevent the fleet from losing control due to excessive acceleration. At the same time, increasing the weights of the two escort ships on both sides helps the fleet to adjust the heading when necessary to ensure navigation safety;

[0151] For example, a three-ship freight formation consisting of 1 main pusher cargo ship and 2 side escort cargo ships transports large goods in a downstream flow of 5 m s. The main pusher cargo ship is located in the downstream direction, with an initial propeller weight of 0.6, which is reduced to 0.4 after adjustment. The initial propeller weights of the two side escort cargo ships are both 0.2, and both are increased to 0.3 after adjustment. The total adjusted weight is 0.4 + 0.3 + 0.3 = 1. By reducing the power of the main pusher cargo ship and enhancing the control ability of the two side escort cargo ships, the formation effectively avoids the risk of losing control due to excessive downstream acceleration and safely delivers the goods to the destination;

[0152] The above quantitative adjustment of weights needs to meet the following constraints:

[0153] 1. The principle of weight conservation, specifically as follows:

[0154] Each time of adjustment, the total increase and decrease of the propeller weights of all ships in the fleet are strictly balanced to ensure that the sum of all weights is always 1 and the total power distribution base remains unchanged;

[0155] 2. The single-ship weight boundary, specifically as follows:

[0156] The propeller weight of a single ship after adjustment needs to satisfy 0 Weight 1;

[0157] For example, when the initial weight of the propeller of a certain ship is 0.2 and it needs to be reduced by 0.3, the actual reduction is 0.2 (the final weight is 0);

[0158] For example, when the initial weight of the propeller of a certain ship is 0.9 and it needs to be increased by 0.2, the actual increase is 0.1 (the final weight is 1).

[0159] Through in-depth extraction and analysis of navigation data, an accurate dynamic allocation feasibility matrix is generated, which comprehensively reflects the efficiency performance of different propellers under the current flow field conditions. This matrix is based on the weighted fusion of the thrust efficiency coefficient and the dynamic demand increment, and combined with the quicksort algorithm to obtain the propeller efficiency priority, providing a scientific basis for subsequent dynamic allocation. Then, the corrected dynamic increment, dynamic margin coefficient, and dynamic allocation feasibility matrix are weighted and fused to generate a dynamic allocation decision parameter set containing the comprehensive performance indicators of each propeller. This process fully considers the actual working conditions of each propeller and the characteristics of the flow field, realizing the refined management of dynamic allocation. Based on the propeller efficiency priority and the dynamic allocation decision parameter set, the power adjustment amount of each propeller is dynamically allocated through the adaptive weight algorithm to generate accurate dynamic allocation instructions, ensuring that the fleet can respond to the flow velocity change with the optimal dynamic configuration in the complex flow field and guaranteeing the navigation safety and efficiency.

[0160] By quantitatively adjusting the weights of ship propellers for different flow field types and flow velocity gradings, following the principles of weight conservation and the boundary constraints of single-ship weights, ensuring that the total power allocation base remains unchanged and the single-ship weights are always within a reasonable range. Under complex flow field conditions such as cross flow, counter current, and forward flow, the system can accurately adjust the weights of the propellers of ships in different positions such as the upwind side, downwind side, front row leading ships, and rear row auxiliary ships according to the specific value and direction of the flow velocity, realizing the effective allocation of power. For example, in low-speed cross flow, slightly adjusting the weights of the propellers of the upwind side and downwind side ships can not only ensure the stable course of the ship but also avoid waste of power. In high-speed cross flow, significantly increasing the weights of the propellers of the upwind side ships can ensure that the fleet can overcome the lateral moment of the strong cross flow and maintain a stable course. This dynamic weight adjustment mechanism based on the characteristics of the flow field significantly improves the course stability and power utilization efficiency of the fleet in the complex flow field, providing a solid power guarantee for the navigation of the combined fleet in complex waters.

[0161] In one case of this embodiment, if the dynamic margin coefficient 1, a hierarchical warning mechanism is triggered, including:

[0162] When 0.8 Dynamic margin coefficient When it is 1, it is determined that there is a critical risk of the remaining power of the main engine in the combined fleet, triggering a first-level warning, and the yellow lights of the combined fleet flash;

[0163] When 0.5 Power margin coefficient When it is 0.8, it is determined that there is a risk of shortage of the remaining power of the main engine in the combined fleet, triggering a second-level warning, and the orange lights of the combined fleet flash;

[0164] When the power margin coefficient When it is 0.5, it is determined that there is a serious shortage risk of the remaining power of the main engine in the combined fleet, triggering a third-level warning, and the red lights of the combined fleet flash.

[0165] In a case of this embodiment, the comprehensive data set is calculated to obtain the time difference between power and flow field, and it is determined whether to execute the power distribution instruction based on the time difference between power and flow field, including:

[0166] Extract the flow field time parameter and power time parameter of the flow field change from the comprehensive data set. Specifically, extract the flow field time parameter of the flow field change from the comprehensive data set. Through the sliding window difference algorithm, the flow velocity data sequence is slid on the time axis according to the set window length (5s), and the mean difference of the flow velocities in adjacent windows is calculated. When the difference exceeds the threshold (1m / s), it is located as the flow velocity mutation point, and the corresponding time is the flow field time parameter; screen out the power distribution instruction records from the comprehensive data set, extract the instruction issuance timestamp, and if the instruction is executed in stages, take the start time of each stage as the power time parameter;

[0167] Calculate the difference between the flow field time parameter and the power time parameter to obtain the time difference between power and flow field. Specifically, subtract the power time parameter from the flow field time parameter to obtain the time difference between power and flow field;

[0168] If the difference between the power and the flow field time ≤ 0, it means that the effective time of the power adjustment lags behind the time of the flow field change, indicating that the power effective time ≥ the flow field influence time (the power adjustment takes effect after or at the same time as the flow field influence), and it is determined not to execute the power distribution instruction and execute the forced trigger instruction;

[0169] If the time difference between power and flow field > 0, it means that the power adjustment can take effect before the flow field mutation affects the ship, indicating that the power effective time < the flow field influence time (the power adjustment takes effect before the flow field change affects the system), and it is determined to execute the power distribution instruction.

[0170] By accurately extracting the flow field time parameters and dynamic time parameters from the comprehensive dataset, and processing the flow velocity data sequence using the sliding window difference algorithm, the flow velocity mutation points are accurately identified and the flow field time parameters are determined. At the same time, the dynamic time parameters are extracted from the dynamic distribution instruction records, ensuring the accuracy of obtaining the time parameters. By calculating the difference between the flow field time parameters and the dynamic time parameters, the time relationship between the dynamic adjustment and the flow field change can be clarified, providing a key basis for whether to execute the dynamic distribution instruction, and improving the navigation safety and stability of the fleet in the complex flow field environment.

[0171] Through the precise determination of the time difference between the power and the flow field, the scientific execution management of the dynamic distribution instruction is realized. When the time difference ≤ 0, the system determines not to execute the dynamic distribution instruction and triggers a forced instruction, effectively avoiding potential risks caused by the lag of dynamic adjustment. When the time difference > 0, the system determines to execute the dynamic distribution instruction, ensuring that the dynamic adjustment can take effect before the flow field change affects the ship, guaranteeing that the fleet can adapt to the flow field change in time and maintain a stable navigation state. This instruction decision-making mechanism based on the time difference significantly improves the navigation efficiency and safety of the fleet in the complex lock environment and reduces the navigation risks caused by improper timing of dynamic adjustment.

[0172] In one case of this embodiment, the hierarchical early warning mechanism further includes:

[0173] At the first-level early warning: Based on the current flow field type (cross flow, counter current and downstream flow), the weight of the high-efficiency thrusters of the ships in the top 50% of the priority is additionally increased by 0.1, and at the same time, the weight of the low-efficiency thrusters of the ships in the bottom 50% of the priority is reduced equally (ensuring that the sum of the weights is 1);

[0174] Temporarily reduce the target speed by 5% - 10%, release the margin by reducing the total power demand, and at the same time maintain the course stability;

[0175] At the second-level early warning: Temporarily increase the weight of the thrust efficiency coefficient to 0.7 (originally 0.6), and reduce the weight of the dynamic demand increment to 0.3;

[0176] For the thrusters of the ships in the backflow side or auxiliary position, the weight upper limit is reduced to 0.2 (originally 0.4), and the weight of the thrusters of the ships in the upstream side or leading ship is increased to 0.7 - 0.8 (not exceeding 1), specifically as follows:

[0177] In the counter current or cross flow scenario: Adjust the weight of the thrusters of the main ships in the upstream side according to the rule upper limit (such as directly increasing by 0.3 in the high-speed cross flow, regardless of whether the initial weight is close to 1, and ensuring that it does not exceed 1), and the weight of the thrusters of the auxiliary ships in the backflow side is reduced to a minimum of 0.05 (not less than 0);

[0178] Downstream scenario: The propeller weight of the front-row ship is reduced to below 0.3, and the propeller weights of the escort ships on both sides are increased to 0.35. The course is controlled by side thrust to reduce the main thrust power consumption;

[0179] The formation changes to a "single-file" formation (reducing the resistance area), and the course of the leading ship is slightly adjusted by 5° - 10° to avoid the strong current direction, reducing the overall water flow resistance;

[0180] The allowable speed is temporarily allowed to be 10% - 20% lower than the target speed, giving priority to ensuring that the power system is not overloaded;

[0181] In the case of a level-three warning: Only 1 ship propeller with the highest efficiency priority is enabled (the weight is directly set to 1, and the remaining propellers are set to 0, satisfying the weight conservation);

[0182] Immediately perform a stop operation, turn off all non-essential loads, and only maintain the operation of navigation and communication equipment;

[0183] Send a distress signal to nearby ports or maritime departments;

[0184] Among them, the level-three warning The level-two warning The level-one warning. High-level warning measures automatically cover low-level ones to avoid conflicts in adjustment strategies (for example, when there is a level-three warning, directly disable inefficient propellers without performing the fine adjustment of the level-one warning).

[0185] Different levels of warnings are accurately triggered through the power margin coefficient and corresponding measures are taken. In the case of a level-one warning, the system adjusts the propeller weights of the ship based on the flow field type, increasing the weights of efficient propellers and reducing those of inefficient ones. At the same time, the target speed is moderately reduced to release the power margin and maintain course stability. In the case of a level-two warning, the thrust efficiency coefficient and the power demand increment weight are adjusted, the weight upper limit of the propellers on the backflow side or auxiliary ships is restricted, the weights of the propellers on the oncoming flow side or the leading ship are increased, and the formation and speed are optimized, giving priority to ensuring that the power system is not overloaded. In the case of a level-three warning, only the most efficient propeller is enabled, the ship is immediately stopped, non-essential loads are turned off, and a distress signal is sent. This hierarchical warning and response measures can maximize the navigation safety of the fleet under different risk levels, reduce the risk of power system overload, and improve the emergency response ability.

[0186] In one case of this embodiment, executing a forced trigger instruction includes:

[0187] Issuing an alarm to remind the operator to manually force an early trigger of power adjustment and enforce the power distribution instruction.

[0188] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A combined fleet intelligent navigation data processing system, characterized in that, Including: A data acquisition unit, which is used to collect the navigation data and lock data of each ship in the combined fleet in real time, fuse the preprocessed navigation data and lock data to obtain a comprehensive data set; A positioning unit, which is used to extract and analyze the features of the comprehensive data set, locate the position and scope of the flow velocity mutation area when the lock is opened and closed, and generate a flow field feature parameter set; An allocation unit, which is used to analyze the flow field feature parameter set and the navigation data to generate a power allocation instruction; A determination unit, which is used to calculate the comprehensive data set to obtain the power and flow field time difference, and determine whether to execute the power allocation instruction based on the power and flow field time difference.

2. The intelligent navigation data processing system for a combined fleet according to claim 1, wherein Fusing the preprocessed navigation data and lock data to obtain a comprehensive data set, including: Analyzing the preprocessed navigation data and lock data to obtain a correlation matrix; Based on the correlation matrix, fusing the preprocessed navigation data and lock data to obtain a fused data set; Performing anomaly processing on the fused data set to obtain a comprehensive data set.

3. The intelligent navigation data processing system for a combined fleet according to claim 2, wherein Performing feature extraction and analysis on the comprehensive data set, locating the position and scope of the flow velocity mutation area when the lock is opened and closed, and generating a flow field feature parameter set, including: Separating the spatio-temporal fluctuation mode of the flow field from the comprehensive data set to generate a dynamic flow field feature vector; Extracting the flow velocity information of each spatial position and time point in the dynamic flow field feature vector, using the central difference method to calculate the modulus of the flow velocity gradient at each spatial point in adjacent spaces and time steps, and setting three gradient thresholds of high, medium, and low; When the modulus of the flow velocity gradient ≥ the high gradient threshold, dividing the flow field space into a high-speed change area, and for the high-speed change area, reducing the grid spacing to one-half of the original spacing; When the low gradient threshold < the modulus of the flow velocity gradient ≤ the medium gradient threshold, dividing the flow field space into a medium-speed change area, and adjusting the grid spacing to three-quarters of the original spacing; When the modulus of the flow velocity gradient ≤ the low gradient threshold, dividing the flow field space into a low-speed change area, and increasing the grid spacing to 2 times the original spacing for sparse processing; Using the Delaunay triangulation algorithm to regenerate grid nodes according to the new grid division, determining the three-dimensional spatial coordinates of the grid nodes, analyzing the connection relationship between the nodes, constructing a grid correlation matrix, recording the information of adjacent grid cells, and finally obtaining a dynamic grid topology structure; Decomposing the dynamic grid topology structure into a steady-state flow velocity field and a transient turbulent flow field, and respectively generating a steady-state flow velocity distribution matrix and a transient vorticity intensity matrix; Comparing the vorticity flow velocity of the transient vorticity intensity matrix with the reference value of the steady-state flow velocity distribution matrix to generate a flow velocity mutation difference field; Normalizing the time change rate of the transient vorticity flow velocity to obtain the ratio of the current flow velocity change rate to the maximum change rate, then normalizing the ship length and the effective length of the lock respectively and dividing them to obtain the ratio of the ship length to the effective length of the lock, and multiplying the ratio of the current flow velocity change rate to the maximum change rate by the ratio of the ship length to the effective length of the lock and adding 1 to form a correction factor; Multiplying the spatially averaged flow velocity difference parameter by the correction factor to obtain a flow velocity mutation difference field.

4. The intelligent navigation data processing system for a combined fleet according to claim 3, characterized in that, Extract and analyze the features of the comprehensive data set, locate the position and range of the flow velocity mutation area during the opening and closing of the ship lock, and generate a set of flow field characteristic parameters, which also includes: Identify the boundaries of the mutation area in the flow velocity mutation difference field to obtain a set of flow velocity mutation boundary coordinates; Calculate the flow velocity gradient components of each grid node in the steady-state flow velocity distribution matrix in the three spatial directions of transverse, longitudinal, and vertical through the central difference method, synthesize these three components into a spatial gradient vector, and calculate the amplitude of this vector to obtain the spatial gradient amplitude; Extract the transient flow velocity data of each node at different times from the dynamic grid topology structure, and calculate the attenuation rate and attenuation multiple based on the transient flow velocity data; Combine the spatial gradient amplitude, attenuation rate, and attenuation multiple to generate a set of turbulent intensity gradient parameters; Fuse the set of flow velocity mutation boundary coordinates and the set of turbulent intensity gradient parameters to generate a set of flow field characteristic parameters.

5. The intelligent navigation data processing system for a combined fleet according to claim 4, wherein Analyze the set of flow field characteristic parameters and the navigation data to generate power distribution instructions, including: Analyze the set of flow field characteristic parameters to obtain the peak flow velocity difference and turbulent influence gradient in the flow velocity mutation area; Divide the draft of the ship by the effective water depth of the ship lock to obtain the water depth ratio, and then divide the opening and closing time of the ship lock by the processing time of the fleet passing through the lock to obtain the time ratio. Multiply the two ratios and add 1 to obtain the ship lock-time correlation correction coefficient; Divide the difference between the peak flow velocity in the flow velocity mutation area and the current ship speed by the maximum steady-state flow velocity and then add 1 to obtain the flow velocity difference-steady-state flow velocity correction coefficient; Multiply the flow velocity gradient-draft relative parameter, the ship lock-time correlation correction coefficient, and the flow velocity difference-steady-state flow velocity correction coefficient to obtain the turbulent influence gradient; Subtract the current ship speed from the peak flow velocity in the flow velocity mutation area to obtain the peak flow velocity difference; Calculate the peak flow velocity difference and the ship draft data to obtain the power increment; Combine and correct the turbulent influence gradient and the power increment to obtain the corrected power increment; Calculate the current output power of the main engine and the corrected power increment to obtain the power margin coefficient. If the power margin coefficient < 1, trigger the hierarchical warning mechanism.

6. The intelligent navigation data processing system for a combined fleet according to claim 5, wherein, Analyze the set of flow field characteristic parameters and the navigation data to generate power distribution instructions, which also includes: Extract the navigation data to generate a power distribution feasibility matrix; Fuse the thrust efficiency coefficient and the power demand increment of the power distribution feasibility matrix according to the weight, and use the weighted summation algorithm to obtain the comprehensive efficiency value of each thruster. In the complex flow field, sort the thrusters from high to low according to the comprehensive efficiency value through the quicksort algorithm to finally obtain the thruster efficiency priority; Fuse the corrected power increment, the power margin coefficient, and the power distribution feasibility matrix to generate a set of power distribution decision parameters including the comprehensive performance indicators of each thruster; According to the thruster efficiency priority and the set of power distribution decision parameters, use the adaptive weight algorithm to dynamically allocate the power adjustment amount of each ship's thruster. The initial weight is allocated according to the thruster efficiency priority of the ship, and then the weights of different ship thrusters are quantitatively adjusted based on the flow field type and flow velocity grading. According to the adjusted weights, allocate the total power adjustment amount to each ship's thruster in proportion to generate the power distribution instructions.

7. An intelligent navigation data processing system for a combined fleet according to claim 5, characterized in that If the power margin coefficient < 1, the hierarchical warning mechanism is triggered, including: When 0.8 ≤ power margin coefficient < 1, it is determined that there is a critical risk of the remaining power of the main engine in the combined fleet, and a first-level warning is triggered, and the yellow lights of the combined fleet flash; When 0.5 ≤ power margin coefficient < 0.8, it is determined that there is a risk of shortage of the remaining power of the main engine in the combined fleet, and a second-level warning is triggered, and the orange lights of the combined fleet flash; When the power margin coefficient < 0.5, it is determined that there is a serious shortage of the remaining power of the main engine in the combined fleet, and a third-level warning is triggered, and the red lights of the combined fleet flash.

8. The intelligent navigation data processing system for a combined fleet according to claim 6, wherein Calculate the comprehensive data set to obtain the time difference between power and flow field, and determine whether to execute the power distribution instruction based on the time difference between power and flow field, including: Extract the flow field time parameter and power time parameter of the flow field change from the comprehensive data set; Calculate the difference between the flow field time parameter and the power time parameter to obtain the time difference between power and flow field; If the difference between the power and the flow field time ≤ 0, it means that the effective time of the power adjustment lags behind the time of the flow field change impact, and it is determined not to execute the power distribution instruction, and the forced trigger instruction is executed; If the time difference between power and flow field > 0, it means that the power adjustment can take effect before the flow field mutation affects the ship, and it is determined to execute the power distribution instruction.

9. The intelligent navigation data processing system for a combined fleet according to claim 1, wherein The hierarchical warning mechanism also includes: At the first-level warning: Based on the current flow field type, the current flow field type is divided into cross flow, counter current and downstream flow. On the basis of the original weight adjustment rule, the weight of the high-efficiency thrusters of the top 50% of the ships in terms of priority is additionally increased by 0.1, and at the same time, the weight of the low-efficiency thrusters of the ships in the last 50% of the priority is reduced equally; Temporarily reduce the target speed by 5% - 10%; At the second-level warning: Temporarily increase the weight of the thrust efficiency coefficient to 0.7 and reduce the weight of the power demand increment to 0.3; For the ship thrusters on the leeward side or in the auxiliary position, the weight upper limit is reduced to 0.2, and the weight of the thrusters on the windward side or the leading ship is increased to 0.7 - 0.

8. Specifically as follows: Counter current or cross flow scenario: Adjust the weight of the thrusters of the main ship on the windward side according to the upper limit of the rule, and the weight of the thrusters of the auxiliary ship on the leeward side is reduced to a minimum of 0.05; Downstream flow scenario: The weight of the thrusters of the front-row ships is reduced below 0.3, and the weight of the thrusters of the two-side escort ships is increased to 0.

35. Control the course through side thrusters to reduce the main thrust power consumption; The formation changes to a "single column" formation, and the course of the leading ship is slightly adjusted by 5° - 10° to avoid the strong flow direction and reduce the overall water flow resistance; The allowable speed is temporarily lower than the target speed by 10% - 20%, and priority is given to ensuring that the power system is not overloaded; At the third-level warning: Only 1 ship thruster with the highest efficiency priority is enabled, and its weight is directly set to 1, and the remaining thrusters are set to 0 to satisfy the weight conservation; Immediately perform a stop operation, turn off all non-essential loads, and only maintain the operation of navigation and communication equipment; Send a distress signal to the nearby port or maritime department; Among them, the third-level warning > the second-level warning > the first-level warning, and the high-level warning measures automatically cover the low-level warning measures.

10. A combined fleet intelligent navigation data processing system according to claim 8, characterized in that, Execute the forced trigger instruction, including: Send an alarm to remind the operator to manually force the early trigger of the power adjustment and enforce the power distribution instruction.

Images