Cross control logic and disturbance self-adaptive route approaching control method and system

Abstract

The invention discloses a cross control logic and disturbance self-adaptive route approaching control method and system, and belongs to the technical field of intelligent ship control. The method comprises the following steps: firstly, generating a refined reference route; then, receiving a cross control instruction containing left / right and front / back instructions, and adjusting the route or speed according to the cross control instruction to generate a new control target; meanwhile, a control strategy adapting to current disturbance is intelligently determined by integrating environment, ship state and load information; and finally, controlling the ship to move based on the new target and strategy, and realizing high-precision route approaching. According to the method, the defects of traditional course keeping logic are overcome, and the control intuition, tracking precision, environmental adaptability and comfort of ship navigation are remarkably improved through visual cross control and intelligent disturbance self-adaption.

Description

Technical Field

[0001] This invention relates to the field of intelligent ship control technology, and in particular to a cross-shaped control logic and disturbance adaptive route approach control method and system. Background Technology

[0002] The core mission of a ship's autopilot system is to precisely control the ship's motion, enabling it to navigate safely, efficiently, and comfortably along a predetermined route. Currently, most mainstream commercial systems are based on autopilots, with their core control logic being course maintenance. This logic typically combines algorithms such as line-of-sight navigation to decompose the macroscopic predetermined route into a series of discrete course target points. The system controls the rudder angle to continuously approach and maintain the ship's bow toward these target course points.

[0003] However, this traditional strategy, which uses "course" as the direct control objective, has significant limitations and struggles to meet the urgent needs for high-precision, highly intelligent navigation in complex waterways (such as inland waterways and congested ports). Specifically, existing technologies suffer from the following main problems: A disconnect between control logic and practical experience: Experienced captains, when maneuvering a ship, do not primarily aim to mechanically maintain a fixed course, but rather to ensure the ship's overall position remains stably "fitted" to the predetermined course. They make overall corrections—"shifting left" or "shifting right"—based on the ship's lateral deviation relative to the course, supplemented by "acceleration" or "deceleration" to adjust the ship's position and cope with encounters. Traditional course-keeping logic cannot effectively simulate this "course-fitting" control mode, which is more in line with human decision-making habits.

[0004] The system's ability to handle environmental disturbances is rudimentary: Existing systems typically treat external disturbances such as wind, waves, and currents as noise that needs to be suppressed, lacking comprehensive perception and intelligent identification capabilities regarding disturbance characteristics (such as direction (heading, lateral, or reverse), and intensity) as well as the ship's own state (such as load condition and speed). This leads to situations where, under strong disturbances, controller parameters remain unchanged, easily resulting in decreased tracking accuracy, frequent steering gear movements, and reduced passenger comfort.

[0005] Insufficient accuracy of control reference data: The raw channel line data used as the input reference for the control law is often coarse, typically providing waypoints only at a scale of a few kilometers. This low-resolution line cannot provide effective geometric constraints for high-precision tracking control, especially in curved or narrow channel areas, making it difficult to meet the needs of fine maneuvering.

[0006] Although advanced algorithms such as model predictive control and reinforcement learning have been introduced into the path tracking field to improve performance, existing technologies have not yet been able to organically integrate the core idea of ​​"course proximity", the control logic that simulates the captain's intuition, intelligent disturbance identification and compensation, and high-precision route generation, so as to build a ship navigation control system with better overall performance, more intelligence, and more in line with actual operating habits. Summary of the Invention

[0007] In view of the above-mentioned shortcomings in the current field of intelligent ship control technology, the present invention provides a cross-shaped control logic and disturbance adaptive route approach control method and system, which can achieve the effect that can be achieved by independent control.

[0008] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions: A cross-shaped control logic and disturbance-adaptive flight path approach control method includes the following steps: Generate a refined reference route for ship tracking; The system receives crosshair control commands and adjusts the refined reference flight path according to the commands to generate a new control target; the crosshair control commands include left / right commands for lateral flight path adjustment and forward / backward commands for longitudinal motion control. Acquire current environmental disturbance information, ship status information, and load information, and determine the control strategy under the current disturbance scenario based on the information; Based on the new control objective and the control strategy, the ship's movement is controlled so that the ship stays close to the route corresponding to the new control objective.

[0009] According to one aspect of the invention, generating a refined reference route for ship tracking includes: Based on the original sparse waypoints, a continuous preliminary reference route is generated using a curve fitting method. Regions with curvature exceeding a preset threshold in the preliminary reference flight path are identified as high maneuverability requirement areas; Within the high maneuverability requirement zone, the trackpoint density is adaptively increased based on the minimum turning radius of the ship predicted by the ship dynamics model to generate the refined reference route.

[0010] According to one aspect of the invention, adjusting the refined reference flight path according to the crosshair control command to generate a new control target includes: In response to the left / right command, the refined reference route is shifted laterally to the left or right to generate a new target tracking route. In response to the aforementioned forward / rear commands, a desired speed target or a desired engine speed target is generated, depending on the type of the ship's propulsion system.

[0011] According to one aspect of the invention, generating targets based on the type of ship propulsion system includes: For electrically driven ships, the desired speed is generated as the control objective. For diesel engine ships, the desired engine speed is generated as the control target.

[0012] According to one aspect of the present invention, determining the control strategy under the current disturbance scenario based on the information specifically includes: Based on rough set theory, the environmental disturbance information, ship state information and load information are used as conditional attributes. A predefined decision table is queried, and a pattern label representing the current disturbance scenario is output. Based on the mode label, one of multiple preset control strategy configurations is selected, wherein the control strategy configuration includes a combination of controller parameter groups and optimization target weight coefficients.

[0013] According to one aspect of the invention, an online optimization step of the control strategy is also included: During the control process, control quality indices, including tracking error, ship motion comfort, and energy consumption indicators, are calculated in real time. When the control quality index continues to fall below the expected performance benchmark of the current pattern label, a correction to the decision table is triggered to update the mapping relationship between the conditional attributes and the pattern label.

[0014] According to one aspect of the invention, the control of ship motion is achieved through a low-level controller, which is one of a model predictive control (MPC) controller, a linear quadratic regulator (LQR) controller, or a proportional-integral-derivative (PID) controller.

[0015] According to one aspect of the present invention, when the underlying controller is a model predictive control (MPC) controller, the control strategy is specifically manifested in dynamically adjusting the weight coefficients and / or constraints of its cost function.

[0016] According to one aspect of the present invention, the specific optimization objectives for controlling ship motion include: minimizing the lateral deviation of the ship from the target route, minimizing the angle between the ship's bow and the tangent direction of the route, and minimizing the ship's rolling motion.

[0017] A cross-shaped control logic and disturbance-adaptive flight path proximity control system includes: The route module generates a refined reference route for ship tracking; The adjustment module receives cross-shaped control commands and adjusts the refined reference flight path according to the commands to generate a new control target; the cross-shaped control commands include left / right commands for lateral flight path adjustment and forward / backward commands for longitudinal motion control; The strategy module acquires current environmental disturbance information, ship status information, and load information, and determines the control strategy under the current disturbance scenario based on the information. The adaptation module, based on the new control objective and the control strategy, controls the ship's movement to make the ship closely follow the route corresponding to the new control objective.

[0018] The advantages of this invention are as follows: By innovating the traditional "heading-keeping" logic into an intuitive "course-following" logic, and pioneering the "cross-shaped" control command, this invention enables the intelligent system to control the vessel to stably follow the course through intuitive "left, right, forward, aft" commands, much like an experienced captain, greatly simplifying operation. Simultaneously, the system possesses intelligent "environmental adaptation" capabilities, automatically identifying disturbances such as wind, waves, and currents and dynamically adjusting control strategies, thereby significantly improving navigation accuracy, safety, and comfort in complex waters and high sea states. By integrating the ship's autopilot control mode, environmental perception, and control core, this invention achieves a paradigm shift from "mechanical execution" to "intelligent following," ultimately making ship navigation safer, more economical, and more intelligent. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the cross-shaped control logic and disturbance-adaptive flight path approach control method and system described in this invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figure 1 As shown, a cross-shaped control logic and disturbance-adaptive flight path approach control method includes the following steps: Generate a refined reference route for ship tracking; The system receives crosshair control commands and adjusts the refined reference flight path according to the commands to generate a new control target; the crosshair control commands include left / right commands for lateral flight path adjustment and forward / backward commands for longitudinal motion control. Acquire current environmental disturbance information, ship status information, and load information, and determine the control strategy under the current disturbance scenario based on the information; Based on the new control objective and the control strategy, the ship's movement is controlled so that the ship stays close to the route corresponding to the new control objective.

[0023] I. An Adaptive Route Refinement Method Based on Channel Geometry and Ship Dynamics The refinement of this invention does not involve indiscriminate high-density sampling of the entire route, but rather adopts an adaptive, on-demand refinement strategy. Its core idea is to provide high-precision guidance in areas where ship maneuverability is difficult (such as sharp bends), while using relatively sparse representations in straight areas to reduce computational burden.

[0024] 1. Initial coarse fitting: First, using the original sparse waypoint set `{P0,P1,...,Pn}`, a preliminary cubic Bézier curve `B0(s)` is constructed. This curve ensures the overall direction and continuity of the route.

[0025] 2. Curvature Analysis and Key Point Extraction: Differential geometric analysis is performed on the initial curve `B0(s)`. The curvature `κ(s)` at each point on the curve is calculated. The curvature `κ` is a key indicator for measuring the degree of curvature, `κ=1 / R` (where R is the radius of curvature). A curvature threshold `κ_th` is set. All regions that satisfy `|κ(s)|>κ_th` are marked as the High-Maneuverability Zone (HMZ).

[0026] 3. Refinement under Ship Dynamics Constraints: Within the HMZ, the refinement density is not arbitrarily set, but determined by the ship's approximate motion prediction model. This invention uses a simplified three-DOF (3-DOF) Nomoto model or a more accurate MMG (Maneuvering Modeling Group, generally 4-DOF, adding heading and roll to the XY plane motion) model as prediction tools.

[0027] Turning Capability Prediction: For each candidate insertion point within the HMZ, the system uses a predictive model to perform back-engineering. Specifically, given the ship's current state (position, heading, speed) and an assumed rudder angle limit `δ_max`, the model can predict the minimum turning radius `R_min` that the ship can complete.

[0028] Refinement Criteria: The predicted `R_min` is compared with the actual radius of curvature of the channel at that point, `R_actual = 1 / |κ(s)|`. If `R_actual` is close to or less than `R_min`, it means that the curve poses a significant challenge to the ship. In this case, the system must significantly increase the density of control points in this area, even increasing the order of the Bézier curve, to accurately characterize the channel geometry and provide a sufficiently smooth and accurate reference trajectory for the underlying controller. Conversely, if `R_actual` is much greater than `R_min`, only moderate densification is needed.

[0029] Dynamic adjustment: This process is iterative. The system continuously inserts new control points within the HMZ and refits the Bézier curve until the local curvature change of the new curve is gradual and its geometry can be "covered" by the ship's dynamic model, meaning that the ship can track the curve without collision within a reasonable range of rudder angles and speeds.

[0030] By combining the channel geometry (curvature) with the ship's dynamic capabilities (minimum turning radius) with this adaptive refinement method, the generated final reference route `P_ref(s)` achieves unprecedented accuracy in key areas while ensuring global smoothness, laying a solid geometric foundation for subsequent high-order intelligent control.

[0031] II. Cross-shaped command parsing and target generation The system receives crosshair commands from the human-machine interface or the upper-level decision-making module. Its design fully considers the technical characteristics of different types of ships, making it widely adaptable. The specific processing logic is as follows: Left / Right Command: Generates a new, refined reference route based on the command. One of the core objectives of subsequent control is to guide the vessel to accurately track the newly generated reference route. This control logic is applicable to vessels of all power types, providing a unified and precise solution for lateral route adjustments.

[0032] Forward / Rearward Commands: Generate corresponding longitudinal control targets based on command requirements, and conduct differentiated adaptation designs to address the technical differences between ships with different power systems. For electrically driven ships, relying on the technical advantages of fast engine response and precise speed control, precise speed control at the level of 0.1 km / h can be achieved. Therefore, the desired speed is used as the control target. For traditional diesel engine ships, considering the technical limitations that high-frequency throttle adjustment may damage the engine, a more fuzzy engine speed is used as the control target.

[0033] III. Deep Coupling Design of the Three Core Controllers and Rough Set Pattern The core of this invention is that the disturbance patterns identified by rough sets are no longer simple suggestions, but rather a direct and dynamic reconfiguration of the internal structure and parameters of the three major controllers: PID, LQR, and MPC, forming a strong coupling relationship of "pattern as configuration".

[0034] 1. Pattern-based reconfiguration of PID controllers: This invention abandons the traditional fixed-parameter PID and designs a pattern-driven reconfigurable PID architecture.

[0035] The pattern determines the structure: the pattern label output by the rough set directly determines the topology of the PID controller. For example, in the "calm waters - high precision" mode, the system activates a dual-loop cascade PID controller (the outer loop follows the course, and the inner loop maintains a stable heading). In the "strong downstream - energy saving" mode, the system may be simplified to a single-loop PD controller, focusing only on lateral offset, because heading stability is naturally better in downstream environments, eliminating the need for a complex inner loop.

[0036] The mode determines the parameters: Each predefined mode is associated with a set of PID parameters `{Kp, Ki, Kd}` optimized through offline simulation and real-ship testing. When the rough set identifies a new mode, the controller parameters smoothly switch to the parameter set of the corresponding mode. For example, the `Kp` value in the "Strong Cross Waves - High Comfort" mode will be significantly lower than that in the "Calm Waters - High Precision" mode to suppress ship rolling caused by frequent rudder inputs.

[0037] 2. Patterned weight matrix of LQR controller: The performance of LQR is entirely determined by the weight matrices `Q` (state weights) and `R` (control weights). This invention completely delegates the design of `Q` and `R` to rough set patterns.

[0038] Pattern-weight mapping table: The system maintains a mapping table that binds each perturbation pattern to a specific `(Q,R)` matrix pair.

[0039] Dynamic weight allocation: For example, in "High Precision" mode, the diagonal elements of the `Q` matrix corresponding to the lateral offset error `e_D` are given extremely high weights, while the elements corresponding to the bow angular velocity `r` have lower weights. In "High Disturbance Immunity" mode, the `Q` matrix significantly increases the weights of `e_ψ` (bow error) and `r`, while the `R` matrix reduces the penalty for the rudder angle `δ`, allowing the controller to use a larger rudder angle to resist disturbances, even if this results in some increased energy consumption. This direct matrix replacement allows LQR to adapt instantly to new control priorities.

[0040] 3. Pattern-based optimization kernel for the MPC controller: MPC's flexibility allows it to make the most of rough set pattern information.

[0041] Pattern-based cost function: Similar to LQR, the cost function weights `w1, w2, w3, w4` of MPC are directly specified by the current pattern.

[0042] Patterned Constraint Set: The constraints of MPC are dynamically adjusted in different modes. For example, in "Emergency Collision Avoidance" mode, the rudder angle rate constraint `|Δδ|` is temporarily relaxed to allow for faster rudder command response. In "High Comfort" mode, the roll acceleration `|ẍ_roll|` is added as a new state constraint.

[0043] Pattern-based prediction time domain: The prediction time domain `N_p` and the control time domain `N_c` can also be determined by the pattern. In the "cruise" mode of a straight channel, a longer `N_p` can be used to optimize global performance; in the "fine maneuver" mode of a narrow channel, a shorter `N_p` is used to ensure real-time performance and rapid response to emergencies.

[0044] IV. Online Evolutionary Mechanism of Rough Set Patterns Based on Control Quality Feedback This invention constructs a complete "perception-decision-execution-evaluation-evolution" closed loop to ensure continuous optimization of system performance.

[0045] 1. Online Control Quality Assessment: In each control cycle (especially the execution cycle of MPC), the system calculates a set of multi-dimensional control quality indicators (Control Quality Index, CQI), including: Accuracy metric: Root mean square (RMS) value of the lateral tracking error `e_D`.

[0046] Stability metrics: standard deviation of bow jitter, frequency of rudder angle movement.

[0047] Comfort metrics: peak and RMS values ​​of roll / pitch acceleration.

[0048] Efficiency metrics: fuel consumption per unit distance or main engine load.

[0049] 2. Model Performance Determination: The system compares the calculated CQI with the performance baseline promised by the current model. If the CQI is consistently (e.g., for 5 consecutive cycles) significantly worse than the baseline (e.g., accuracy error exceeds the limit by 30%), the current model is determined to be "failed" under this operating condition.

[0050] 3. Evolution of Online Mode: Case Recording: The system records this "failure" event as a new case in the rough set case library, including complete conditional attributes (disturbance, state, load) and decision results (selected mode) as well as CQI.

[0051] Rule Correction: When a sufficient number of "failure" cases accumulate under the same condition attribute combination, the system will automatically trigger the rule correction algorithm. This algorithm analyzes the case library to find a better decision attribute. For example, the original rule might be "Strong Waves + Half Load -> High Precision Mode," but after multiple failures, the algorithm will find that "Strong Waves + Half Load -> High Disturbance Immunity Mode" has a better CQI. Therefore, the system will automatically update the decision table, correcting the decision result of this condition combination to "High Disturbance Immunity Mode."

[0052] New rule verification: The revised rule will enter a short "trial period" during which its CQI will be closely monitored. Only after passing the verification will the new rule be officially adopted.

[0053] Advantages of this invention: Through this mechanism, the control system of this invention possesses powerful online learning and self-evolution capabilities. It is no longer a static system dependent on prior knowledge, but an intelligent agent capable of continuously learning from its own control experience and constantly optimizing its decision-making logic. Thus, in any complex and ever-changing navigation environment, it can autonomously find and apply the most suitable control strategy to achieve safe, efficient, and comfortable intelligent navigation.

[0054] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cross-shaped control logic and disturbance-adaptive flight path approach control method, characterized in that, Includes the following steps: Generate a refined reference route for ship tracking; The system receives crosshair control commands and adjusts the refined reference flight path according to the commands to generate a new control target; the crosshair control commands include left / right commands for lateral flight path adjustment and forward / backward commands for longitudinal motion control. Acquire current environmental disturbance information, ship status information, and load information, and determine the control strategy under the current disturbance scenario based on the information; Based on the new control objective and the control strategy, the ship's movement is controlled so that the ship stays close to the route corresponding to the new control objective.

2. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 1, characterized in that, The generation of a refined reference route for ship tracking includes: Based on the original sparse waypoints, a continuous preliminary reference route is generated using a curve fitting method. Regions with curvature exceeding a preset threshold in the preliminary reference flight path are identified as high maneuverability requirement areas; Within the high maneuverability requirement zone, the trackpoint density is adaptively increased based on the minimum turning radius of the ship predicted by the ship dynamics model to generate the refined reference route.

3. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 1, characterized in that, The adjustment of the refined reference flight path according to the cross-shaped control command to generate a new control target includes: In response to the left / right command, the refined reference route is shifted laterally to the left or right to generate a new target tracking route. In response to the aforementioned forward / rear commands, a desired speed target or a desired engine speed target is generated, depending on the type of the ship's propulsion system.

4. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 3, characterized in that, The generation of targets based on the type of ship propulsion system includes: For electrically driven ships, the desired speed is generated as the control objective. For diesel engine ships, the desired engine speed is generated as the control target.

5. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 1, characterized in that, The specific steps for determining the control strategy under the current disturbance scenario based on the information are as follows: Based on rough set theory, the environmental disturbance information, ship state information and load information are used as conditional attributes. A predefined decision table is queried, and a pattern label representing the current disturbance scenario is output. Based on the mode label, one of multiple preset control strategy configurations is selected, wherein the control strategy configuration includes a combination of controller parameter groups and optimization target weight coefficients.

6. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 5, characterized in that, It also includes online optimization steps for the control strategy: During the control process, control quality indices, including tracking error, ship motion comfort, and energy consumption indicators, are calculated in real time. When the control quality index continues to fall below the expected performance benchmark of the current pattern label, a correction to the decision table is triggered to update the mapping relationship between the conditional attributes and the pattern label.

7. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 1, characterized in that, The control of the ship's motion is achieved through a bottom-level controller, which is one of a model predictive control (MPC) controller, a linear quadratic regulator (LQR) controller, or a proportional-integral-derivative (PID) controller.

8. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 7, characterized in that, When the underlying controller is a model predictive control (MPC) controller, the control strategy is specifically manifested in dynamically adjusting the weight coefficients and / or constraints of its cost function.

9. The cross-shaped control logic and disturbance-adaptive flight path approach control method according to claim 1, characterized in that, The specific optimization objectives for controlling ship motion include: minimizing the lateral deviation of the ship from the target route, minimizing the angle between the ship's bow and the tangent of the route, and minimizing the ship's rolling motion.

10. A cross-shaped control logic and disturbance-adaptive flight path approach control system, characterized in that, A cross-shaped control logic and disturbance-adaptive flight path approach control method based on any one of claims 1 to 9 includes: The route module generates a refined reference route for ship tracking; The adjustment module receives cross-shaped control commands and adjusts the refined reference flight path according to the commands to generate a new control target; the cross-shaped control commands include left / right commands for lateral flight path adjustment and forward / backward commands for longitudinal motion control; The strategy module acquires current environmental disturbance information, ship status information, and load information, and determines the control strategy under the current disturbance scenario based on the information. The adaptation module, based on the new control objective and the control strategy, controls the ship's movement to make the ship closely follow the route corresponding to the new control objective.

Images