Method for controlling one-key truss crane of geotextile laying ship

Abstract

The invention discloses a one-key truss crane control method for a geotextile laying ship. The method comprises the steps that S1, an operator inputs operation parameters through a control terminal; s2, generating a preliminary operation path according to the input parameters; s3, whether the states of the truss crane and the lifting appliance are normal or not is checked, and operation is started after confirmation; s4, after operation is started, the movement of the truss crane is made according to the operation path, and the position of the lifting appliance is fed back through the real-time position feedback technology; s5, the positions and torque states of the truss crane and the lifting appliance are monitored in real time, whether deviation occurs in path execution or not is judged, if deviation occurs, the path is locally optimized through dynamic adjustment, and the current working state is fed back to an operator through voice and visual prompt; and S6, after laying of all the interlocking blocks is completed, the truss crane is automatically reset to the initial position, and an operation report is output. Automatic control and intelligent adjustment technologies are combined, and the operation convenience, safety and efficiency of the geotextile laying ship in the interlocking block laying operation are improved.

Description

Technical Field

[0001] The invention relates to a one-button truss crane control method for a laying vessel. Background Art

[0002] Laying vessels are specialized vessels widely used in underwater engineering projects, primarily for laying interlocking blocks to form stable protective or support structures. Laying interlocking blocks typically requires high precision and efficiency. Traditional truss cranes rely primarily on manual control, which is not only complex and labor-intensive, but also prone to misalignment due to control errors, impacting both operational quality and project safety.

[0003] In order to solve the above problems, a one-button truss crane control method for a laying vessel is provided. Summary of the Invention

[0004] The purpose of the present invention is to overcome the existing defects and provide a one-button truss crane control method for a laying vessel, which combines automatic control and intelligent adjustment technology to improve the operational convenience, safety and efficiency of the laying vessel in the interlocking block laying operation.

[0005] The technical solution to achieve the above purpose is:

[0006] A one-button truss crane control method for a laying vessel, comprising:

[0007] Step S1, the operator inputs the operation parameters through the control terminal;

[0008] Step S2, generating a preliminary operation path according to the input parameters;

[0009] Step S3, check whether the truss crane and the spreader are in normal condition, and start the operation after confirmation;

[0010] Step S4: After the operation is started, the movement of the truss crane is controlled according to the operation path, and the position of the spreader is fed back using real-time position feedback technology;

[0011] Step S5: Real-time monitoring of the position and torque of the truss crane and spreader to determine whether there are any deviations in the path execution. If there are any deviations, the path is locally optimized through dynamic adjustment, and the current operation status is fed back to the operator through voice and visual prompts;

[0012] Step S6: After all interlocking blocks are laid, the truss crane is automatically reset to its initial position and a work report is output;

[0013] Step S7: Using a dynamic obstacle detection module to scan the operation area in real time, update obstacle location information, and perform global optimization of the path using a genetic algorithm;

[0014] Step S8: Adjust the spreader motion strategy in real time based on the reinforcement learning model, and optimize the path tracking accuracy according to historical operation data and real-time feedback.

[0015] Preferably, in step S1, the operation parameters include the scope of the operation area, the laying spacing of the interlocking blocks and the initial laying order.

[0016] Preferably, in step S2, generating a preliminary operation path according to the input parameters includes:

[0017] Step S21, generating two-dimensional coordinates of all points to be laid according to the interlocking block laying spacing and the scope of the working area;

[0018] Step S22, sorting all the laying points using a gridding method and connecting the points according to the shortest path principle;

[0019] Step S23, detecting obstacles in the path and adjusting the movement trajectory of the spreader;

[0020] Step S24: Using a gradient descent algorithm to optimize the path curve, and finally generating a motion path of the truss crane.

[0021] Preferably, in step S21, a gridded coordinate point set is established according to the boundary of the working area and the interlocking block laying spacing. The working area is defined as a two-dimensional plane, which is recorded as a rectangular area, that is:

[0022] [x min ,x max ]×[y min ,y max ];

[0023] Assuming the spacing between laying points is d, the coordinate set P of the grid points is:

[0024]

[0025] Where i and j represent the row index and column index of the grid point on the two-dimensional plane, corresponding to the horizontal and vertical grid numbers respectively. The value range is determined by the size of the paving area and the grid spacing d. Indicates rounding down, L x and L y represent the horizontal and vertical path lengths respectively.

[0026] Preferably, in step S22, in the generated paving point set P, a paving sequence with the shortest total moving distance is found through path sorting optimization, that is, the point set P is connected according to the paving sequence, and the total length L0 of the initial path is:

[0027]

[0028] In the formula, (x k ,y k ) is the coordinate of the kth laying point, (x k+1 ,y k+1 ) is the coordinate of the k+1th laying point, and N is the total number of laying points;

[0029] Use the dynamic programming algorithm of the traveling salesman problem to optimize the path sequence, and set the distance between any two points in the point set P to be:

[0030]

[0031] That is, the optimization goal is to minimize the total path length L:

[0032]

[0033] Where, σ is the arrangement of the laying points, S N is the set of all possible permutations;

[0034] The recursive equation of dynamic programming is:

[0035]

[0036] Where D(S,i) represents the shortest path starting from set S and finally reaching i.

[0037] Preferably, in step S23, during path planning, if there are obstacles in the working area, the path needs to be adjusted:

[0038] Define the obstacle area O, then:

[0039] O={(x,y)|(xx c ) 2 +(yy c ) 2 ≤r 2};

[0040] In the formula, (x c ,y c ) is the center point of the obstacle, r is the radius of the obstacle;

[0041] If a point (x, y)∈O in the path, then reinsert the replacement point so that the path bypasses the obstacle area and dynamically adjusts the path point set to P';

[0042] Connect the point set P according to the laying order ′ , the adjusted path length L ′ for:

[0043]

[0044] Where M is the total number of path points after obstacle avoidance.

[0045] Preferably, in step S24, if the preliminary path contains a sharp turning point or a discontinuous path, path smoothing optimization is required, and the optimization objectives include the path curvature change rate K and minimizing the total path length L;

[0046] The calculation formula of the path curvature change rate K is as follows:

[0047]

[0048] In the formula, (x k ,y k ) is the kth path point;

[0049] The constraint condition is that the path point (x k ,y k ) must remain within the grid and not affect the coverage of the laying points, that is:

[0050] x min ≤x k ≤x max ;

[0051] y min ≤y k ≤y max ;

[0052] Use the gradient descent method to optimize and minimize the total path length L, and calculate the path points (x k ,y k ) for iterative update, where λ is the weight coefficient and the update formula is:

[0053]

[0054] Where α is the learning rate, t is the number of iterations, and t is set to 1000 times, which can be adjusted according to actual conditions.

[0055] Preferably, in step S3, whether the truss crane and the spreader are in normal condition is determined by whether the relevant sensor signals exceed the limit value. If so, it is abnormal; wherein the sensor signals include but are not limited to signals generated by the encoder, motor current and proximity limit switch;

[0056] In step S5, during the execution of the path, the current position (x real ,y real ), and with the planned path point (x k ,y k ) for comparison, if the error value:

[0057]

[0058] ε is the allowable deviation, which triggers path adjustment:

[0059]

[0060] Dynamic adjustments include:

[0061] If a dynamic obstacle is detected intruding into the path, the emergency obstacle avoidance mode is triggered, a smooth detour path is generated using the Bezier curve, and the path is updated to the truss crane control system in real time;

[0062] In step S6, the outputted work report includes but is not limited to laying accuracy, time consumption and fault records.

[0063] Preferably, in step S7, the dynamic obstacle detection module obtains the obstacle position in real time through the laser radar and visual sensor fusion technology, and updates the obstacle set U as:

[0064] U(t)={(x,y,t)|(xx c (t)) 2 +(yy c (t)) 2 ≤r(t) 2};

[0065] In the formula, (x c (t),y c (t)) is the coordinate of the center point of the dynamic obstacle, r(t) is the radius of the dynamic obstacle, both of which change with time t;

[0066] Path optimization uses genetic algorithms, including:

[0067] Step S71, initialize the population and encode the current path point sequence into a chromosome;

[0068] Chromosome encoding uses real number encoding, and each gene corresponds to the coordinates of the path point (x k ,y k ), the population size is set to 100-500, and the mutation probability is 0.01-0.1;

[0069] Step S72, calculating the fitness function F;

[0070]

[0071] Where L is the total length of the path, C is the rate of change of path curvature, and λ is the weight coefficient;

[0072] Step S73: Generate a new generation of population through selection, crossover and mutation operations, and iterate and optimize until the termination condition is met.

[0073] Preferably, in step S8, the reinforcement learning model takes the current position deviation, torque state and environmental state of the spreader as input and outputs the path tracking correction; the model is trained by the Q-learning algorithm, and the reward function R is defined as:

[0074] R = -(α·deviation+β·energy consumption+γ·time);

[0075] Where α is the weight emphasizing position accuracy, β is the weight of motor energy consumption, and γ is the weight of optimizing operation time.

[0076] The beneficial effects of the present invention are as follows: the present invention significantly reduces the complexity of manual operation through path planning and one-button control; based on the optimized path planning algorithm, the lifting path is shortened, time consumption is reduced, and work efficiency is improved; obstacle avoidance is taken into account in path planning, which reduces operational risks and ensures safety; real-time adjustment and position feedback functions ensure the precise laying of interlocking blocks and enhance accuracy; and the present invention can be flexibly applied to interlocking block laying tasks of different sizes and shapes, and has strong practicality and promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 This is a flow chart of a one-button truss crane control method for a laying vessel according to the present invention;

[0078] Figure 2 It is a specific flow chart of generating a preliminary operation path according to input parameters in the present invention;

[0079] Figure 3 It is a specific flow chart of the global optimization of paths by genetic algorithm in the present invention. DETAILED DESCRIPTION

[0080] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0081] The present invention will be further described below with reference to the accompanying drawings.

[0082] like Figure 1 As shown, a one-button truss crane control method for a laying vessel includes:

[0083] In step S1, the operator inputs operation parameters through the control terminal.

[0084] In an embodiment, the operation parameters include the scope of the operation area, the interlocking block laying spacing, and the initial laying sequence.

[0085] Step S2: Generate a preliminary operation path according to the input parameters.

[0086] like Figure 2 As shown, a preliminary job path is generated based on the input parameters, including:

[0087] Step S21: Generate the two-dimensional coordinates of all points to be laid according to the laying spacing of the interlocking blocks and the scope of the working area.

[0088] In the embodiment, a grid coordinate point set is established based on the boundary of the working area and the interlocking block laying spacing. The working area is defined as a two-dimensional plane, which is recorded as a rectangular area, that is:

[0089] [x min ,x max ]×[y min ,y max ];

[0090] Assuming the spacing between laying points is d, the coordinate set P of the grid points is:

[0091]

[0092] Where i and j represent the row index and column index of the grid point on the two-dimensional plane, corresponding to the horizontal and vertical grid numbers respectively. The value range is determined by the size of the paving area and the grid spacing d. Indicates rounding down, L x and L y represent the horizontal and vertical path lengths respectively.

[0093] Step S22: sort all the laying points using a gridding method and connect the points according to the shortest path principle.

[0094] In this embodiment, in the generated paving point set P, the paving sequence with the shortest total moving distance is found through path sorting optimization. That is, the point set P is connected according to the paving sequence, and the total length L0 of the initial path is:

[0095]

[0096] In the formula, (x k ,y k ) is the coordinate of the kth laying point, (x k+1 ,y k+1) is the coordinate of the k+1th laying point, and N is the total number of laying points;

[0097] Use the dynamic programming algorithm of the traveling salesman problem to optimize the path sequence, and set the distance between any two points in the point set P to be:

[0098]

[0099] That is, the optimization goal is to minimize the total path length L:

[0100]

[0101] Where, σ is the arrangement of the laying points, S N is the set of all possible permutations;

[0102] The recursive equation of dynamic programming is:

[0103]

[0104] Where D(S,i) represents the shortest path starting from set S and finally reaching i.

[0105] Step S23: Detect obstacles in the path and adjust the movement trajectory of the spreader to ensure that there is no conflict in the path.

[0106] In the embodiment, during path planning, if there are obstacles in the working area, the path needs to be adjusted:

[0107] Define the obstacle area O, then:

[0108] O={(x,y)|(xx c ) 2 +(yy c ) 2 ≤r 2};

[0109] In the formula, (x c ,y c ) is the center point of the obstacle, r is the radius of the obstacle;

[0110] If a point (x, y)∈O in the path, then reinsert the replacement point so that the path bypasses the obstacle area and dynamically adjusts the path point set to P ′ ;

[0111] Connect the point set P according to the laying order ′ , the adjusted path length L ′ for:

[0112]

[0113] Where M is the total number of path points after obstacle avoidance.

[0114] Step S24 , using a gradient descent algorithm to optimize the path curve, ultimately generating a motion path for the truss crane, thereby reducing sudden stops or mutations during the movement of the truss crane.

[0115] In the embodiment, if the preliminary path includes a sharp turning point or a discontinuous path, path smoothing optimization is required, and its optimization objectives include the path curvature change rate K and minimizing the total path length L;

[0116] The calculation formula of the path curvature change rate K is as follows:

[0117]

[0118] In the formula, (x k ,y k ) is the kth path point;

[0119] The constraint condition is that the path point (x k ,y k ) must remain within the grid and not affect the coverage of the laying points, that is:

[0120] x min ≤x k ≤x max ;

[0121] y min ≤y k ≤y max ;

[0122] Use the gradient descent method to optimize and minimize the total path length L, and calculate the path points (x k ,y k ) for iterative update, where λ is the weight coefficient and the update formula is:

[0123]

[0124] Where α is the learning rate, t is the number of iterations, and t is set to 1000 times, which can be adjusted according to actual conditions.

[0125] Step S3: Check whether the truss crane and the spreader are in normal condition, and start the operation after confirmation.

[0126] In the embodiment, whether the status of the truss crane and the sling is normal is determined by whether the relevant sensor signals exceed the limit value. If so, it is abnormal; wherein the sensor signals include but are not limited to signals generated by the encoder, motor current and proximity limit switch.

[0127] Step S4: After the operation is started, the movement of the truss crane is controlled according to the operation path, and the position of the spreader is fed back using real-time position feedback technology.

[0128] Step S5 monitors the position and torque status of the truss crane and the spreader in real time to determine whether there is any deviation in the path execution. If there is a deviation, the path is locally optimized through dynamic adjustment, and the current operation status is fed back to the operator through voice and visual prompts.

[0129] In the embodiment, the positions of the truss crane and the sling are determined by installing an encoder to calculate the rope length, and the torque is determined by the torque provided by the motor.

[0130] In the embodiment, during the path execution process, the current position (x real ,y real ), and with the planned path point (x k ,y k ) for comparison, if the error value:

[0131]

[0132] ε is the allowable deviation, which triggers path adjustment:

[0133]

[0134] In an embodiment, dynamic adjustment includes:

[0135] If a dynamic obstacle is detected intruding into the path, the emergency obstacle avoidance mode is triggered, a smooth detour path is generated through the Bezier curve, and the path is updated to the truss crane control system in real time.

[0136] Step S6: After all interlocking blocks are laid, the truss crane is automatically reset to its initial position and an operation report is output.

[0137] In the embodiment, the truss crane is automatically reset to its initial position to confirm that all the interlocking blocks have been laid.

[0138] In the embodiment, the output work report includes but is not limited to paving accuracy, time consumption and fault records, providing data support for subsequent optimization.

[0139] Step S7: Use the dynamic obstacle detection module to scan the operation area in real time, update the obstacle position information, and perform global optimization of the path through the genetic algorithm.

[0140] In this embodiment, the dynamic obstacle detection module obtains the obstacle position in real time through the laser radar and visual sensor fusion technology, and updates the obstacle set U as:

[0141] U(t)={(x,y,t)|(xx c (t)) 2 +(yy c (t)) 2≤r(t) 2};

[0142] In the formula, (x c (t),y c (t)) is the coordinate of the center point of the dynamic obstacle, r(t) is the radius of the dynamic obstacle, both of which change with time t;

[0143] A multi-line laser radar and high-definition camera are installed on the top of the truss, and a real-time environment map is constructed using SLAM technology. The obstacle position is updated every 0.1 seconds.

[0144] like Figure 3 As shown in Figure 2, path optimization uses a genetic algorithm, which specifically includes:

[0145] Step S71, initialize the population and encode the current path point sequence into a chromosome;

[0146] Chromosome encoding uses real number encoding, and each gene corresponds to the coordinates of the path point (x k ,y k ), the population size is set to 100-500, and the mutation probability is 0.01-0.1;

[0147] Step S72, calculating the fitness function F;

[0148]

[0149] Where L is the total length of the path, C is the rate of change of path curvature, and λ is the weight coefficient;

[0150] Step S73: Generate a new generation of population through selection, crossover and mutation operations, and iterate and optimize until the termination condition is met.

[0151] In the embodiment, the number of iterations is set to 500 generations, the crossover rate is set to 0.8, and the roulette wheel method is used for the selection operation, and a smooth global path with the optimal length is finally generated.

[0152] Step S8: Adjust the spreader motion strategy in real time based on the reinforcement learning model, and optimize the path tracking accuracy according to historical operation data and real-time feedback.

[0153] In this embodiment, the reinforcement learning model takes the current position deviation, torque state, and environmental state of the spreader as input and outputs a path tracking correction. The model is trained using the Q-learning algorithm to achieve millisecond-level response. The reward function R is defined as:

[0154] R = -(α·deviation+β·energy consumption+γ·time);

[0155] Where α is the weight emphasizing position accuracy, β is the weight of motor energy consumption, and γ is the weight of optimizing operation time.

[0156] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A one-button truss crane control method for a laying vessel, characterized in that: include: Step S1, the operator inputs the operation parameters through the control terminal; Step S2, generating a preliminary operation path according to the input parameters; Step S3, check whether the truss crane and the spreader are in normal condition, and start the operation after confirmation; Step S4: After the operation is started, the movement of the truss crane is controlled according to the operation path, and the position of the spreader is fed back using real-time position feedback technology; Step S5: Real-time monitoring of the position and torque of the truss crane and spreader to determine whether there are any deviations in the path execution. If there are any deviations, the path is locally optimized through dynamic adjustment, and the current operation status is fed back to the operator through voice and visual prompts; Step S6: After all interlocking blocks are laid, the truss crane is automatically reset to its initial position and a work report is output; Step S7: Using a dynamic obstacle detection module to scan the operation area in real time, update obstacle location information, and perform global optimization of the path using a genetic algorithm; Step S8: Adjust the spreader motion strategy in real time based on the reinforcement learning model, and optimize the path tracking accuracy according to historical operation data and real-time feedback.

2. A one-button truss crane control method for a laying vessel according to claim 1, characterized in that: In step S1, the operation parameters include the scope of the operation area, the laying spacing of the interlocking blocks, and the initial laying order.

3. The one-button truss crane control method for a laying vessel according to claim 1, characterized in that: In step S2, a preliminary operation path is generated according to the input parameters, including: Step S21, generating two-dimensional coordinates of all points to be laid according to the interlocking block laying spacing and the scope of the working area; Step S22, sorting all the laying points using a gridding method and connecting the points according to the shortest path principle; Step S23, detecting obstacles in the path and adjusting the movement trajectory of the spreader; Step S24: Using a gradient descent algorithm to optimize the path curve, and finally generating a motion path of the truss crane.

4. A one-button truss crane control method for a laying vessel according to claim 3, characterized in that: In step S21, a gridded coordinate point set is established based on the boundary of the working area and the interlocking block laying spacing. The working area is defined as a two-dimensional plane, which is recorded as a rectangular area, namely: [x min ,x max ]×[and min ,and max ]; Assuming the spacing between laying points is d, the coordinate set P of the grid points is: Where i and j represent the row index and column index of the grid point on the two-dimensional plane, corresponding to the horizontal and vertical grid numbers respectively. The value range is determined by the size of the paving area and the grid spacing d. Indicates rounding down, L x and L y represent the horizontal and vertical path lengths respectively.

5. The one-button truss crane control method for a laying vessel according to claim 3, characterized in that: In step S22, in the generated paving point set P, the paving sequence with the shortest total moving distance is found through path sorting optimization. That is, the point set P is connected according to the paving sequence, and the total length of the initial path L0 is: In the formula, (x k ,y k ) is the coordinate of the kth laying point, (x k+1 ,y k+1 ) is the coordinate of the k+1th laying point, and N is the total number of laying points; Use the dynamic programming algorithm of the traveling salesman problem to optimize the path sequence, and set the distance between any two points in the point set P to be: That is, the optimization goal is to minimize the total path length L: Where, σ is the arrangement of the laying points, S N is the set of all possible permutations; The recursive equation of dynamic programming is: Where D(S,i) represents the shortest path starting from set S and finally reaching i.

6. The one-button truss crane control method for a laying vessel according to claim 3, characterized in that: In step S23, during path planning, if there are obstacles in the working area, the path needs to be adjusted: Define the obstacle area O, then: O={(x,y)|(xx c ) 2 +(yy c ) 2 ≤r 2 }; In the formula, (x c ,y c ) is the center point of the obstacle, r is the radius of the obstacle; If a point (x, y)∈O in the path, then reinsert the replacement point so that the path bypasses the obstacle area and dynamically adjusts the path point set to P′; Connect the point set P′ according to the laying order, and the adjusted path length L′ is: Where M is the total number of path points after obstacle avoidance.

7. The one-button truss crane control method for a laying vessel according to claim 3, characterized in that: In step S24, if the preliminary path contains sharp turning points or a discontinuous path, path smoothing optimization is required, and the optimization objectives include the path curvature change rate K and minimizing the total path length L; The calculation formula of the path curvature change rate K is as follows: In the formula, (x k ,y k ) is the kth path point; The constraint condition is that the path point (x k ,y k ) must remain within the grid and not affect the coverage of the laying points, that is: x min ≤x k ≤x max ; and min ≤y k ≤y max ; Use the gradient descent method to optimize and minimize the total path length L, and calculate the path points (x k ,y k ) for iterative update, where λ is the weight coefficient and the update formula is: Where α is the learning rate, t is the number of iterations, and t is set to 1000 times, which can be adjusted according to actual conditions.

8. The one-button truss crane control method for a laying vessel according to claim 1, characterized in that: In step S3, whether the truss crane and the spreader are in normal condition is determined by whether the relevant sensor signals exceed the limit value. If so, it is abnormal; wherein the sensor signals include but are not limited to signals generated by the encoder, motor current and proximity limit switch; In step S5, during the execution of the path, the current position (x real ,y real ), and with the planned path point (x k ,y k ) for comparison, if the error value: ε is the allowable deviation, which triggers path adjustment: Dynamic adjustments include: If a dynamic obstacle is detected intruding into the path, the emergency obstacle avoidance mode is triggered, a smooth detour path is generated using the Bezier curve, and the path is updated to the truss crane control system in real time; In step S6, the outputted work report includes but is not limited to laying accuracy, time consumption and fault records.

9. The one-button truss crane control method for a laying vessel according to claim 1, characterized in that: In step S7, the dynamic obstacle detection module obtains the obstacle position in real time through the laser radar and visual sensor fusion technology, and updates the obstacle set U as: U(t)={(x,y,t)|(x-x c (t)) 2 +(y-y c (t)) 2 ≤r(t) 2 }; In the formula, (x c (t),y c (t)) is the coordinate of the center point of the dynamic obstacle, r(t) is the radius of the dynamic obstacle, both of which change with time t; Path optimization uses genetic algorithms, including: Step S71, initialize the population and encode the current path point sequence into a chromosome; Chromosome encoding uses real number encoding, and each gene corresponds to the coordinates of the path point (x k ,y k ), the population size is set to 100-500, and the mutation probability is 0.01-0.1; Step S72, calculating the fitness function F; Where L is the total length of the path, C is the rate of change of path curvature, and λ is the weight coefficient; Step S73: Generate a new generation of population through selection, crossover and mutation operations, and iterate and optimize until the termination condition is met.

10. The one-button truss crane control method for a laying vessel according to claim 1, characterized in that: In step S8, the reinforcement learning model takes the current position deviation, torque state, and environmental state of the spreader as input and outputs a path tracking correction. The model is trained using the Q-learning algorithm, and the reward function R is defined as: R = -(α·deviation+β·energy consumption+γ·time); Where α is the weight emphasizing position accuracy, β is the weight of motor energy consumption, and γ is the weight of optimizing operation time.

Images