An automatic ship driving control method and system based on artificial intelligence

Abstract

The present invention discloses an automatic ship driving control method and system based on artificial intelligence, which relates to the technical field of ship control. The driving control steps are as follows: S1. Obtain the current two-dimensional coordinate position of the ship based on the satellite map, obtain the wind direction data at the current two-dimensional coordinate position in the shipping lane based on the meteorological bureau, and obtain the shipping lane state to construct a data set; S2. Establish a ship driving model, combine the real-time wind direction data to predict the position and attitude changes of the ship in the future shipping lane, and through the control algorithm, calculate the ship displacement control amount in advance to offset the deviation caused by the wind direction factor. The present invention ensures that the ship can return to the correct shipping lane in a timely and accurate manner, greatly improves the navigation accuracy of the ship in a complex environment, realizes the refined adjustment of the ship navigation parameters, avoids unnecessary energy waste, avoids deviating from the target shipping lane, and increases the risks of collision and grounding.

Description

Technical Field

[0001] The present invention relates to the technical field of ship control, and specifically to an automatic ship driving control method and system based on artificial intelligence. Background Art

[0002] At present, with the continuous development of the global shipping industry, due to its large transportation capacity and low cost, ship transportation occupies a crucial position in the fields of cargo transportation and people's livelihood trade, undertaking 95% of the global crude oil transportation and 99% of the iron ore transportation volume. However, the traditional ship driving mode faces severe challenges in terms of safety, efficiency, and cost. To effectively solve the above problems, ship autonomous driving technology has emerged and become a research hotspot. Early ship autonomous driving mainly relied on traditional control methods such as PID control and fuzzy control. PID control calculates the output signal of the steering gear by the deviation between the ship's target course and the actual course and related change amounts to maintain the ship's course, and combines the consideration of the change of wind direction during the ship's travel to adjust and change the position of the ship during travel. However, due to the large inertia of the ship, when the autonomous driving system adjusts the control strategy according to the change of wind direction, the ship cannot respond immediately. Especially in strong wind weather, after the wind direction suddenly changes, the ship may take a long time to adjust to the new course. During this period, the ship may deviate from the target route, increasing the risks of collision, grounding, etc., and also increasing the power energy consumption of the ship to a certain extent. In view of this, we propose an automatic ship driving control method and system based on artificial intelligence. Summary of the Invention

[0003] To solve the above technical problems, an automatic ship driving control method and system based on artificial intelligence are provided. This technical solution solves the problems that the ship cannot respond immediately, the ship will deviate from the target route, and the risks of collision and grounding are increased.

[0004] To achieve the above object, the technical solution adopted by the present invention is: an automatic ship driving control method based on artificial intelligence, and the driving control steps are as follows:

[0005] S1. Obtain the current two-dimensional coordinate position of the ship based on the satellite map, obtain the wind direction data at the current two-dimensional coordinate position in the route based on the meteorological bureau, and obtain the ship route state to construct a data set;

[0006] S2. Establish a ship travel model, combine the real-time wind direction data to predict the position and attitude changes of the ship in the future route, and through the control algorithm, calculate the ship displacement control amount in advance to offset the deviation caused by the wind direction factor;

[0007] S3. When encountering a sudden strong wind causing deviation from the route, generate a route correction path based on the A* algorithm;

[0008] S4. Based on the current inertial state, steering condition and route safety of the ship, select the best correction path from the generated correction routes and sail to the route;

[0009] S5. Based on the optimal control algorithm, taking the ship displacement control amount, energy consumption and structural stress as optimization objectives, establish the objective function and constraint conditions. Based on the particle swarm optimization algorithm, when the ship corrects its course to sail to the route, find a set of optimal control parameters and control the ship's sailing based on the optimal control parameters to reduce energy consumption.

[0010] Preferably, the steps for establishing the ship sailing model in step S2 are as follows:

[0011] Collect the basic parameters of the ship itself and the wind direction data on the route;

[0012] Based on Newton's laws of motion and fluid mechanics, construct the kinematic and dynamic equations of the ship, and in the form of a state space model, establish a mathematical relationship between the ship's motion state variables and input variables to form the framework of the ship sailing model;

[0013] Input the real-time wind direction data into the model, calculate the force and moment of the wind on the ship, and combine with the marine environmental factors to establish the coupling of the hydrodynamic model and the ship motion model;

[0014] Based on the historical navigation data and test data, verify the established model, adjust the model parameters, and complete the construction of the ship sailing model.

[0015] Preferably, the steps for calculating the ship displacement control amount in advance through the control algorithm in step S2 are as follows:

[0016] Based on the obtained ship navigation state and wind direction data, after filtering, input them into the ship sailing model;

[0017] Combine the wind direction data to predict the future position and attitude of the ship, and determine the offset caused by the wind direction;

[0018] Use the control algorithm to calculate the displacement control amount, input the displacement control amount into the ship control system, and control the ship displacement to offset the offset;

[0019] Monitor the error between the actual motion state and the ideal state, dynamically adjust the algorithm parameters, and form a closed-loop control.

[0020] Preferably, the steps for generating the route correction path by the A* algorithm in step S3 are as follows:

[0021] Collect the real-time state data of the ship, and the state data includes position and speed information;

[0022] Obtain the surrounding wind direction, sea condition environment data and geographical data, grid the navigation area on the electronic nautical chart, and determine the starting and target nodes;

[0023] Set a heuristic function that incorporates environmental factors, set the parameters of the A* algorithm, use the A* algorithm to search for a path. Starting from the starting node, expand the nodes according to the total cost, avoid areas with poor sea conditions, and backtrack to generate a corrected path after finding the target node.

[0024] Preferably, the grid processing of the route area determines the side length of the grid on the electronic nautical chart based on the ship size, and performs grid processing according to the side length.

[0025] After the grid is completed, determine the starting node and the target node. The starting node is the grid center corresponding to the actual position of the ship after deviating from the current route. The target node is selected according to the original route plan, and the grid center corresponding to the next key position point that the ship should reach without being affected is selected.

[0026] The steps of the A* algorithm to search for a path are as follows: Search for a path from the starting node, calculate the total cost of the adjacent passable nodes of the starting node, select the node with the minimum total cost as the next node to be expanded. When expanding the node, preferentially search in areas with good sea conditions, expand in sequence, continuously update the cost and parent node information of the node, construct a search tree. During the search process, judge in real time whether the target node is reached. If the target node is reached, then by backtracking the parent node information, starting from the target node, gradually return to the starting node along the parent node pointer, and generate a preliminary route correction path from the current position of the ship to the target position. The path includes several different options.

[0027] Preferably, the specific steps to screen out the best corrected path in step S4 are as follows:

[0028] Collect real-time ship data and preprocess the collected data.

[0029] Based on the preprocessed data, calculate the quantization values of the evaluation indicators, including:

[0030] Inertial state index, obtained by calculating the matching degree between the corrected path and the current inertial state of the ship.

[0031] Steering situation index, by calculating whether the requirement of the corrected path for the ship's steering ability is within the operable range; if the steering angle exceeds the maximum steering ability of the ship, the score is low; if the steering requirement conforms to the steering performance of the ship and the ship can complete the steering operation smoothly, the score is high.

[0032] Route safety index, obtained by evaluating the safety of the corrected path during navigation, considering factors such as the distance from the path to obstacles, the probability of passing through dangerous areas, and the influence of meteorological conditions on the path.

[0033] Preferably, the specific steps to screen out the best corrected path further include:

[0034] Based on the hierarchical analysis method, the weight values corresponding to the inertia state index, the turning situation index and the route safety index are determined;

[0035] For each generated revised path, the scores of each evaluation indicator on the path are calculated separately. A scoring system is used to score based on the degree of compliance between the path and the indicator. The comprehensive score of each indicator in the current revised path is calculated using the weighted average formula to obtain the quantitative evaluation results of each path.

[0036] Compare the comprehensive scores of all correction paths and select the path with the highest score as the best correction path;

[0037] If there are multiple paths with the same score, the scores of the paths on the key indicators are further compared, and the path with the higher key indicator score is prioritized to determine the best correction path;

[0038] The optimal corrected path information is sent to the ship's navigation and control system to guide the ship safely back to the planned route.

[0039] Preferably, the objective function established in step S5 includes: displacement control target, energy consumption target and structural stress target;

[0040] The displacement control target compares the actual position and heading of the ship with the target value. The smaller the difference, the closer it is to the ideal route. This difference is quantified as the displacement control cost.

[0041] The energy consumption target is achieved by establishing an energy consumption calculation model based on the characteristics of the ship's power system, taking into account the impact of propulsion force on energy consumption, and calculating the total energy consumption cost during navigation;

[0042] The structural stress target estimates the stress on the ship structure by analyzing the changes in rudder angle and propulsion force. The higher the stress level, the greater the corresponding stress cost.

[0043] The three objectives are added together according to their proportional weights to form an objective function that comprehensively measures the ship's operating performance.

[0044] Preferably, the step of finding the optimal control parameters in step S5 is:

[0045] The control parameters of the ship's rudder angle and main engine speed are input into the established objective function. The particle fitness is evaluated by simulating energy consumption. The inertia weight factor is introduced to find the ship's driving control parameters corresponding to the optimal particle position. During the ship's driving process, the ship's driving is controlled based on the found ship's driving control parameters.

[0046] An automatic ship driving control system based on artificial intelligence, the driving control system includes:

[0047] A data acquisition module configured to collect data on the ship's driving status and wind direction during the route;

[0048] A wind direction offset control module is configured to calculate the ship displacement control amount in advance to offset the offset caused by wind direction factors;

[0049] a course correction module configured to generate a plurality of course correction paths;

[0050] a screening module configured to screen out an optimal correction path;

[0051] The control optimization module is configured to obtain optimal ship driving parameters for driving.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention combines the ship's driving model to predict future position and attitude changes, and calculates the displacement control amount in advance, which can effectively offset the deviation caused by wind direction, so that the ship's navigation trajectory is more in line with the planned route. Even if the ship deviates from the route due to sudden strong winds, the A* algorithm is used to quickly generate a correction path, and then the best correction route is determined through screening to ensure that the ship can return to the correct route in a timely and accurate manner. This greatly improves the navigation accuracy of the ship in complex environments, and realizes the refined adjustment of the ship's navigation parameters, avoiding unnecessary energy waste, avoiding deviation from the target route, and increasing the risk of collision and grounding. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a flow chart of the driving control steps of the present invention.

[0055] Figure 2 It is a framework diagram of the driving control system of the present invention. DETAILED DESCRIPTION

[0056] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0057] Reference Figure 1 As shown, an automatic ship driving control method based on artificial intelligence, the driving control steps are:

[0058] S1. Obtain the current two-dimensional coordinate position of the ship based on the satellite map, obtain the wind direction data of the current two-dimensional coordinate position within the route based on the meteorological bureau, and obtain the ship route status to construct a data set;

[0059] S2. Build a ship driving model and combine it with real-time wind direction data to predict the ship's position and attitude changes in the future route. Use a control algorithm to calculate the ship's displacement control value in advance to offset the offset caused by wind direction factors.

[0060] S3: When encountering sudden strong winds that cause the route to deviate, the route correction path is generated based on the A* algorithm;

[0061] S4. Based on the current inertia state, steering situation and route safety of the ship, the optimal correction path is selected from the generated correction route and the ship is driven into the route;

[0062] S5. Based on the optimization control algorithm, the ship displacement control amount, energy consumption and structural stress are used as optimization targets. The objective function and constraint conditions are established. Based on the particle swarm optimization algorithm, a set of optimal control parameters are found when the ship corrects its course. The ship is controlled based on the optimal control parameters to reduce energy consumption.

[0063] By combining satellite maps and meteorological bureau data, this application can accurately obtain the current two-dimensional coordinate position of the ship and real-time wind direction data, and build a data set. This enables the ship to have a clear and accurate understanding of its own position and surrounding environment, providing a solid foundation for subsequent driving control and helping to plan and respond to various situations in advance;

[0064] By building a ship navigation model and combining it with real-time wind direction data, we can predict the ship's position and attitude changes on the future route. Through the control algorithm, we can calculate the ship's displacement control value in advance to offset the wind direction deviation. This active control method can effectively reduce the impact of wind direction on ship navigation, improve navigation accuracy and stability, and reduce navigation errors caused by wind direction uncertainty.

[0065] When encountering sudden strong winds that cause the ship to deviate from the route, the A* algorithm is used to generate a route correction path, which can quickly find the best path from the current position to return to the original route or adjust to a new safe route, providing the ship with an effective response strategy in emergency situations and enhancing the ship's ability to respond to emergencies; the generated correction route is screened based on the ship's current inertia state, steering situation and route safety, ensuring that the selected best correction path can not only return the ship to the route, but also fully consider the actual operating status and safety factors of the ship, avoiding danger or damage to the ship due to unreasonable correction paths; the ship's displacement control amount, energy consumption and structural stress are used as optimization targets, the objective function and constraints are established, and the optimal control parameters are found based on the particle swarm optimization algorithm. This method can effectively reduce energy consumption and operating costs in the process of controlling the ship's navigation, while also protecting the ship's structure and extending the ship's service life, thus achieving multi-objective optimization of navigation control.

[0066] The steps for establishing the ship driving model in step S2 are:

[0067] Collect basic parameters of the ship and wind direction data during the route;

[0068] Based on Newton's laws of motion and fluid mechanics, kinematic and dynamic equations of the ship are constructed. In the form of a state-space model, a mathematical relationship is established between the ship's motion state variables and input variables, forming a framework for the ship's motion model.

[0069] Real-time wind direction data is input into the model to calculate the forces and moments exerted on the ship by the wind. Combining with ocean environmental factors, a coupling of the hydrodynamic model and the ship motion model is established.

[0070] Based on historical navigation data and test data, the established model is verified and the model parameters are adjusted to complete the construction of the ship motion model.

[0071] The model constructed in this application through Newton's laws of motion and fluid mechanics can accurately depict the ship's motion state in a complex ocean environment. After coupling data such as real-time wind direction, waves, and ocean currents into the model, the ship can predict in advance the impact of harsh environments on its own motion. For example, before strong winds or huge waves arrive, the crew can, based on the calculation results of the model, plan a safer route in advance, adjust the ship's speed and heading, effectively avoid dangerous areas, and reduce the probability of accidents such as collisions and capsizes, providing a solid guarantee for the safety of the ship and personnel. The framework of the ship motion model established based on the state-space model can clearly present the relationship between the ship's motion state variables and input variables, and different navigation plans can be simulated using the model to analyze the most fuel-efficient and shortest-time-consuming routes and operation strategies.

[0072] In step S2, the steps for calculating the ship displacement control amount in advance through the control algorithm are as follows:

[0073] Based on the obtained ship navigation state and wind direction data, after filtering, it is input into the ship motion model.

[0074] Combined with the wind direction data, predict the future position and attitude of the ship, and determine the offset caused by the wind direction.

[0075] Use the control algorithm to calculate the displacement control amount, input the displacement control amount into the ship control system, and control the ship displacement to offset the offset.

[0076] Monitor the error between the actual motion state and the ideal state, and dynamically adjust the algorithm parameters to form a closed-loop control.

[0077] The specific calculation steps are as follows:

[0078] Definition of state vector:

[0079]

[0080] Where x k , y k are the ship's position coordinates, ψ k is the ship's heading angle, are the longitudinal and transverse speeds of the ship, r k is the ship’s bow angular velocity;

[0081] The wind direction data vector is:

[0082]

[0083] in is the wind speed, is the wind direction angle, is the windward angle;

[0084] The prediction formula based on the state space model is:

[0085]

[0086] Where A is the system matrix, B is the control input matrix, E is the interference input matrix, d k The disturbance vector including the effect of wind direction;

[0087] The formula for calculating the offset caused by wind direction is:

[0088]

[0089] Among them G w is the wind direction effect gain matrix, f w (·) is the wind direction influence function, including:

[0090]

[0091] The formula for predicting future position and posture is:

[0092]

[0093] Where N is the length of the prediction time domain;

[0094] The formula for calculating the displacement control amount is:

[0095] u control =[δ rudder ,T prop ] T

[0096] where δ rudder is the rudder angle control quantity, T prop is the propeller thrust control quantity, u control In order to calculate the displacement control amount, the position control amount is input into the ship control system to accurately control the ship's displacement to offset the deviation caused by wind direction in real time.

[0097] In step S3, the A* algorithm generates the corrected path as follows:

[0098] Collect real-time status data of the ship, where the status data includes position and speed information;

[0099] Obtain the surrounding wind direction, sea condition environment data and geographical data, perform grid processing on the navigation area on the electronic nautical chart, and determine the starting and target nodes;

[0100] Set a heuristic function that incorporates environmental factors, set the parameters of the A* algorithm, use the A* algorithm to search for a path, start from the starting node, expand the nodes according to the total cost, avoid areas with poor sea conditions, and backtrack to generate a corrected path after finding the target node.

[0101] This application can dynamically generate a corrected path according to the current actual situation by collecting real-time status data of the ship and the surrounding wind direction, sea condition environment data and geographical data. Whether the ship encounters sudden bad sea conditions or the original route is no longer optimal due to a change in wind direction, the A* algorithm can timely adjust according to the latest information, enabling the ship to always sail along the path most suitable for the current environment, improving the ship's ability to cope with complex and changeable marine environments. The A* algorithm itself is efficient. After performing grid processing on the navigation area on the electronic nautical chart, starting from the starting node and expanding the nodes according to the total cost, it can quickly search for the target node and generate a corrected path. This search method avoids blind search and greatly reduces the search space and time.

[0102] The grid processing of the route area determines the side length of the grid on the electronic nautical chart based on the ship size and performs grid processing according to the side length;

[0103] After the grid processing is completed, the starting node and the target node are determined. The starting node is the grid center corresponding to the actual position of the ship after deviating from the current route, and the target node is selected according to the original route plan as the grid center corresponding to the next key position point that the ship should reach without being affected;

[0104] Among them, the A* algorithm is specifically expressed as:

[0105] The navigation area is divided into grids, and the position of each grid is a node. Among them, the current position of the ship is the starting node, and the destination position is the target node. All grid nodes form a set. To measure the cost of traveling from the starting node to each node, we define three costs: the actual cost of traveling from the starting node to a certain node is called the actual cost; the estimated cost of traveling from a certain node to the target node is called the heuristic estimated cost; the sum of the actual cost and the heuristic estimated cost is the total cost. Each node also has a sea condition score, and the larger the value, the better the sea condition of the area. At the same time, a sea condition weight coefficient is set to reflect the influence degree of the sea condition on the path planning;

[0106] When calculating the cost of a node, first calculate the basic heuristic cost without considering the sea state, which is represented by the straight-line distance from the node to the target node. Incorporate the sea state factor. The better the sea state, the lower the estimated cost from this node to the target node. Add the actual cost from the starting node to this node and the estimated cost considering the sea state to obtain the total cost of this node;

[0107] After starting the search for the path, first create two lists. One open list stores the nodes waiting to be expanded, and one closed list stores the nodes that have been expanded. Put the starting node into the open list. At this time, set the actual cost of the starting node to 0, and calculate its basic heuristic cost and the total cost considering the sea state;

[0108] Find the node with the minimum total cost from the open list as the current node to be expanded. Move it from the open list to the closed list. Then find the adjacent and passable nodes of this node, and calculate the movement cost from the current node to the adjacent node. The better the sea state, the lower the movement cost. If the newly calculated actual cost of the adjacent node is less than the previously recorded actual cost of this node, update the actual cost and total cost of this node, and set the current node as its parent node;

[0109] Continuously repeat the above operations of selecting and expanding nodes. During this process, real-time judge whether the currently expanded node is the target node. If it is, it means that the path has been found;

[0110] After finding the target node, trace back to the starting node step by step through the parent node information recorded by the target node, and thus generate a preliminary route correction path from the current position of the ship to the target position;

[0111] We find the nodes in the open list and the closed list whose total costs are not much different from the total cost of the target node. Using these nodes as the end points, trace back to generate different paths in the same way, and these paths constitute several route correction path options.

[0112] The specific steps for screening the best correction path in step S4 are as follows:

[0113] Collect the real-time data of the ship and preprocess the collected data;

[0114] Based on the preprocessed data, calculate the quantization values of the evaluation indexes, including:

[0115] The inertial state index, which is obtained by calculating the matching degree between the correction path and the current inertial state of the ship;

[0116] Steering condition indicator: This indicator calculates whether the ship's steering capability requirements for the corrected path are within the operational range. If the steering angle exceeds the ship's maximum steering capability, the score is low; if the steering requirement meets the ship's steering performance and the ship can successfully complete the steering operation, the score is high.

[0117] The route safety index is obtained by evaluating the safety of the corrected path during navigation, taking into account the distance between the path and obstacles, the probability of passing through dangerous areas, and the impact of meteorological conditions on the path.

[0118] This application calculates the route safety index and comprehensively considers factors such as the distance between the corrected path and obstacles, the probability of passing through dangerous areas, and meteorological conditions. It can effectively avoid potential dangers, reduce collision risks, ensure the safety of ships and personnel, and improve the reliability and stability of navigation. It considers the inertia state index to match the corrected path with the current inertia state of the ship, which can reduce unnecessary acceleration, deceleration and steering, reduce energy consumption, improve navigation efficiency, save time and fuel costs, evaluate the steering situation index, and ensure that the requirements of the ship's steering ability for the corrected path are within the operational range, allowing the ship to turn smoothly and avoid route deviation or navigation delays due to steering difficulties.

[0119] The specific steps of selecting the best correction path further include:

[0120] Based on the hierarchical analysis method, the weight values corresponding to the inertia state index, the turning situation index and the route safety index are determined;

[0121] For each generated revised path, the scores of each evaluation indicator on the path are calculated separately. A scoring system is used to score based on the degree of compliance between the path and the indicator. The comprehensive score of each indicator in the current revised path is calculated using the weighted average formula to obtain the quantitative evaluation results of each path.

[0122] Compare the comprehensive scores of all correction paths and select the path with the highest score as the best correction path;

[0123] If there are multiple paths with the same score, the scores of the paths on the key indicators are further compared, and the path with the higher key indicator score is prioritized to determine the best correction path;

[0124] The optimal corrected path information is sent to the ship's navigation and control system to guide the ship safely back to the planned route.

[0125] This application determines the weight values of each evaluation index through the analytic hierarchy process, which can more scientifically and accurately reflect the relative importance of different indexes in path evaluation. For each modified path, the scores of each evaluation index are calculated and the comprehensive score value is obtained through weighted average. This quantitative evaluation method can comprehensively and objectively measure the overall quality of the path, avoiding the one-sidedness of single-index evaluation and making the evaluation results more accurate and reliable.

[0126] When there are multiple paths with the same score, further compare the scores of the paths on the key indexes, and preferentially select the path with a higher score on the key indexes. This ensures that when multiple choices seem similar, decisions can be made based on the factors that are more critical to navigation, further improving the rationality and optimization degree of the screening results, and making the selected best path better meet the actual navigation needs.

[0127] The objective function established in step S5 includes: displacement control objective, energy consumption objective and structural stress objective;

[0128] The displacement control objective quantifies the difference between the actual position and heading of the ship and the target value. The smaller the difference, the closer it is to the ideal route, and this difference is quantified as the displacement control cost.

[0129] The energy consumption objective calculates the total energy consumption cost during navigation by establishing an energy consumption calculation model based on the characteristics of the ship's power system and considering the influence of the propulsion force on energy consumption.

[0130] The structural stress objective estimates the stress borne by the ship's structure by analyzing the changes in rudder angle and propulsion force. The higher the stress level, the greater the corresponding stress cost.

[0131] The three objectives are added together according to the proportional weights to form an objective function for comprehensively measuring the operation effect of the ship.

[0132] The specific steps for constructing the objective function are as follows:

[0133] Displacement control objective

[0134]

[0135] where \(u = [\delta,n]\) T is the control parameter vector, \(x(t),y(t),\psi(t)\) are the actual abscissa position of the ship at time \(t\), the actual ordinate position of the ship at time \(t\) and the actual heading angle of the ship at time \(t\) respectively,

[0136] x d (t),y d (t),\(\psi\) d (t) are the abscissa position of the ship's target at time \(t\), the ordinate position of the ship's target at time \(t\) and the heading angle of the ship's target at time \(t\) respectively, \(w\)x , w y , w ψ are the horizontal position error weight, vertical position error weight, and course angle error weight, respectively;

[0137] Energy consumption target

[0138]

[0139] where ρ is the water density, g is the acceleration due to gravity, C T (v x ) is the drag coefficient, related to the ship speed, K P is the propeller coefficient;

[0140] Structural stress target

[0141]

[0142] where T = ρ · n 2 · D 4 · K T (J) is the propulsive force, D is the propeller diameter, K T (J) is the propulsive force coefficient, related to the advance coefficient J, K δ , K T is the stress weight coefficient;

[0143] Calculation of the comprehensive objective function

[0144] J(u) = λ d · J d (u) + λ e · J e (u) + λ s · J s (u)

[0145] where λ d , λ e , λ s are the weight coefficients of each target, satisfying λ d + λ e + λ s = 1, and J(u) is the objective function obtained by comprehensive calculation.

[0146] The steps to find the optimal control parameters in step S5 are as follows:

[0147] During the ship's navigation, precise and efficient control is crucial to ensure the ship arrives at the destination safely and economically. The rudder angle and main engine speed of the ship are two key control parameters, and their reasonable adjustment can significantly affect various performances of the ship's navigation trajectory, speed, and energy consumption;

[0148] Establish an objective function that comprehensively considers multiple factors. This objective function covers multiple dimensions including displacement control objectives, energy consumption objectives, and structural stress objectives. The displacement control objective quantifies the difference between the actual position and heading of the ship and the target values as the displacement control cost by comparing them, thereby ensuring that the ship sails along the ideal route as much as possible. The energy consumption objective establishes an energy consumption calculation model based on the characteristics of the ship's power system, fully considering factors such as the magnitude of the propulsion force on energy consumption, and accurately calculates the total energy consumption cost during navigation, so as to achieve energy conservation and consumption reduction. The structural stress objective estimates the stress borne by the ship's structure by analyzing the changes in the rudder angle and propulsion force. The higher the stress level, the greater the corresponding stress cost, thereby ensuring the safety of the ship's structure;

[0149] Input the two control parameters of the ship's rudder angle and main engine speed into the above-established objective function. With the help of advanced simulation technology, simulate the navigation state of the ship under different combinations of control parameters. During the simulation process, focus on calculating the energy consumption of the ship, which is used as an important basis for evaluating the fitness of particles. Here, introduce the concept in the particle swarm optimization algorithm, and regard different combinations of control parameters as different positions of particles. Each particle has its corresponding fitness. The higher the fitness, the better the performance of the combination of control parameters corresponding to the particle position in meeting the objective function;

[0150] Introduce an inertia weight factor. The inertia weight factor can balance the global search and local search capabilities of particles. At the initial stage of the search, a larger inertia weight factor makes particles more inclined to global search, and can explore the possible solution space in a wider range, avoiding the algorithm from converging to the local optimal solution prematurely. As the search progresses, the inertia weight factor gradually decreases, and the local search ability of particles is enhanced, and they can perform more refined searches near the current optimal solution, increasing the probability of finding the global optimal solution. By continuously iteratively updating the positions and velocities of particles and continuously calculating the fitness of particles, finally find the optimal particle position. The control parameters such as the ship's rudder angle and main engine speed corresponding to this optimal particle position are the parameter combinations that can make the ship operate best under the current objective function and constraints.

[0151] An automatic ship driving control system based on artificial intelligence. The driving control system includes:

[0152] A data acquisition module configured to collect the ship's driving state and wind direction data on the route;

[0153] A wind direction cancellation control module configured to calculate the ship's displacement control amount in advance to cancel the offset caused by the wind direction factor;

[0154] A route correction module configured to generate multiple route correction paths;

[0155] A screening module configured to screen out the best correction path;

[0156] The control optimization module is configured to obtain the optimal ship driving parameters for driving.

[0157] By combining the ship driving model to predict future position and attitude changes and calculating the displacement control amount in advance, it can effectively offset the deviation caused by the wind direction, making the ship's navigation trajectory more conform to the predetermined route. Even when the ship deviates from the route due to sudden strong winds, the A* algorithm is used to quickly generate a correction path, and then the best correction route is determined through screening to ensure that the ship can return to the correct route in a timely and accurate manner, greatly improving the navigation accuracy of the ship in complex environments, realizing the refined adjustment of the ship's navigation parameters, avoiding unnecessary energy waste, avoiding deviating from the target route, and increasing the risks of collision and grounding.

[0158] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. An automatic ship driving control method based on artificial intelligence, characterized in that, The driving control steps are: S1. Obtain the current two-dimensional coordinate position of the ship based on the satellite map, obtain the wind direction data of the current two-dimensional coordinate position within the route based on the meteorological bureau, and obtain the ship route status to construct a data set; S2. Build a ship driving model and combine it with real-time wind direction data to predict the ship's position and attitude changes in the future route. Use a control algorithm to calculate the ship's displacement control value in advance to offset the offset caused by wind direction factors. S3: When encountering sudden strong winds that cause the route to deviate, the route correction path is generated based on the A* algorithm; S4. Based on the current inertia state, steering situation and route safety of the ship, the optimal correction path is selected from the generated correction route and the ship is driven into the route; S5. Based on the optimization control algorithm, the displacement control amount, energy consumption and structural stress of the ship are taken as optimization targets. The objective function and constraint conditions are established. Based on the particle swarm optimization algorithm, a set of optimal control parameters is found when the ship corrects its course. The ship is controlled based on the optimal control parameters to reduce energy consumption. The specific steps for selecting the best correction path in step S4 are: Collect real-time data from ships and pre-process the collected data; Based on the preprocessed data, the quantitative values of the evaluation indicators are calculated, including: The inertial state index is obtained by calculating the degree of match between the corrected path and the current inertial state of the ship; Steering condition indicator: This indicator calculates whether the steering capability requirements of the ship for the corrected path are within the operational range. If the steering angle exceeds the maximum steering capability of the ship, the score is low; if the steering requirement meets the ship's steering performance and the ship can successfully complete the steering operation, the score is high. The route safety index is obtained by evaluating the safety of the corrected route during navigation, taking into account the distance between the route and obstacles, the probability of passing through dangerous areas, and the impact of weather conditions on the route; The specific steps of selecting the best correction path further include: Based on the hierarchical analysis method, the weight values corresponding to the inertia state index, the turning situation index and the route safety index are determined; For each generated revised path, the scores of each evaluation indicator on the path are calculated separately. A scoring system is used to score based on the degree of compliance between the path and the indicator. The comprehensive score of each indicator in the current revised path is calculated using the weighted average formula to obtain the quantitative evaluation results of each path. Compare the comprehensive scores of all correction paths and select the path with the highest score as the best correction path; If there are multiple paths with the same score, the scores of the paths on the key indicators are further compared, and the path with the higher key indicator score is prioritized to determine the best correction path; Send the optimal corrected path information to the ship's navigation and control system to guide the ship safely back to the planned route; The objective functions established in step S5 include: displacement control target, energy consumption target and structural stress target; The displacement control target compares the actual position and heading of the ship with the target value. The smaller the difference, the closer it is to the ideal route. This difference is quantified as the displacement control cost. The energy consumption target is achieved by establishing an energy consumption calculation model based on the characteristics of the ship's power system, taking into account the impact of propulsion force on energy consumption, and calculating the total energy consumption cost during navigation; The structural stress target estimates the stress borne by the ship's structure by analyzing the changes in rudder angle and propulsive force. The higher the stress level, the greater the corresponding stress cost. The three targets are added together according to proportional weights to form an objective function for comprehensively measuring the ship's operating effect.

2. The automatic ship driving control method based on artificial intelligence according to claim 1, characterized in that The steps for establishing the ship motion model in step S2 are as follows: Collect the basic parameters of the ship itself and the wind direction data along the route. Based on Newton's laws of motion and fluid mechanics, construct the kinematic and dynamic equations of the ship. In the form of a state-space model, establish a mathematical relationship between the ship's motion state variables and input variables to form the framework of the ship motion model. Input the real-time wind direction data into the model, calculate the forces and moments exerted on the ship by the wind, and combine with ocean environmental factors to establish the coupling of the hydrodynamic model and the ship motion model. Based on historical navigation data and experimental data, verify the established model, adjust the model parameters, and complete the construction of the ship motion model.

3. The automatic ship driving control method based on artificial intelligence according to claim 1, wherein The steps for calculating the ship displacement control amount in advance through the control algorithm in step S2 are as follows: Based on the obtained ship navigation state and wind direction data, after filtering, input it into the ship motion model. Combine the wind direction data to predict the future position and attitude of the ship, and determine the offset caused by the wind direction. Use the control algorithm to calculate the displacement control amount, input the displacement control amount into the ship control system to control the ship displacement and offset the deviation. Monitor the error between the actual motion state and the ideal state, and dynamically adjust the algorithm parameters to form a closed-loop control.

4. The automatic ship driving control method based on artificial intelligence according to claim 1, characterized in that The steps for generating the route correction path by the A* algorithm in step S3 are as follows: Collect the real-time state data of the ship, and the state data includes position and speed information. Obtain the surrounding wind direction, sea condition environment data and geographical data, perform grid processing on the navigation area on the electronic nautical chart, and determine the starting and target nodes. Set the heuristic function that integrates environmental factors, set the parameters of the A* algorithm, use the A* algorithm to search for the path, start from the starting node, expand the nodes according to the total cost, avoid the areas with poor sea conditions, and generate the correction path by backtracking after finding the target node.

5. The automatic ship driving control method based on artificial intelligence according to claim 4, characterized in that The grid processing of the route area determines the side length of the grid on the electronic nautical chart based on the ship size, and perform grid processing according to the side length. After the grid processing is completed, determine the starting node and the target node. The starting node is the grid center corresponding to the actual position of the ship after deviating from the current route, and the target node is selected according to the original route plan as the grid center corresponding to the next key position point that the ship should reach without being affected. The steps for the A* algorithm to search for the path are as follows: Search for the path from the starting node, calculate the total cost of the adjacent passable nodes of the starting node, select the node with the minimum total cost as the next node to be expanded. When expanding the nodes, preferentially search in the areas with good sea conditions, expand in turn, continuously update the cost and parent node information of the nodes, construct the search tree. During the search process, judge in real time whether the target node is reached. If the target node is reached, then through backtracking the parent node information, starting from the target node, gradually return to the starting node along the parent node pointer to generate a preliminary route correction path from the current position of the ship to the target position, and the path includes several different options.

6. The automatic ship driving control method based on artificial intelligence according to claim 1, wherein The steps for finding the optimal control parameters in step S5 are as follows: Input the control parameters of the ship's rudder angle and main engine speed into the established objective function, evaluate the fitness of particles for energy consumption through simulation calculation, introduce an inertia weight factor, find the ship's driving control parameters corresponding to the optimal particle position, and control the ship's driving based on the found ship's driving control parameters during the ship's driving process.

7. An automatic ship driving control system based on artificial intelligence, which is applied to an automatic ship driving control method based on artificial intelligence according to any one of the above claims 1-6, characterized in that, Including: A data acquisition module configured to collect the ship's driving state and wind direction data in the route; A wind direction offset control module configured to calculate the ship displacement control amount in advance to offset the offset generated by the wind direction factor; A route correction module configured to generate multiple route correction paths; A screening module configured to screen out the best correction path; A control optimization module configured to obtain the best ship driving parameters for driving.

Images