Shipborne servo system control method and system based on multi-dimensional data analysis

Abstract

The application discloses a shipborne servo system control method and system based on multidimensional data analysis and belongs to the technical field of servo control. The system comprises a multidimensional data sensing module, a servo prediction analysis module, a pointing demonstration planning module and a posture adjustment control module. The multidimensional data sensing module is used for collecting historical records, three-dimensional models, monitoring data and environmental characteristic data of a ship. A virtual simulation model is built and dynamically mapped. The servo prediction analysis module uses the historical records to train the virtual simulation model, predicts a pointing demonstration track of an antenna pedestal, constructs a two-dimensional curved track surface and analyzes a signal strength distribution situation. The pointing demonstration planning module analyzes an angular velocity change curve according to the pointing demonstration track, divides an abnormal area, establishes an adjustment area for the abnormal area, establishes an adjustment scheme, replans a new pointing demonstration track in combination with the signal strength distribution situation and controls the antenna pedestal on the ship to perform posture adjustment according to the new pointing demonstration track.

Description

Technical Field

[0001] This invention relates to the field of servo control technology, specifically to a shipborne servo system control method and system based on multi-dimensional data analysis. Background Technology

[0002] In the fields of shipboard satellite communication, radar detection, and various wireless information transmission, shipboard servo systems undertake the core task of driving antennas or sensors to accurately point and stably track remote targets. Their performance directly determines the reliability of communication links and the quality of data transmission in complex marine environments. With the increasing demand for high-bandwidth, low-latency, and continuous communication from ocean voyages, marine monitoring, and offshore operations, unprecedented challenges are being placed on the ability of shipboard servo systems to maintain high-performance tracking under dynamic and harsh sea conditions.

[0003] Currently, mainstream shipborne servo control technologies are typically based on feedback control theory and inertial stabilization platforms. However, under multi-dimensional, strongly coupled, and nonlinear marine disturbances, existing solutions still have inherent limitations and key problems. For example: 1. Mismatch between control models and reality caused by singular environmental perception and modeling. Traditional methods rely on sensor feedback, treating disturbances as simple interferences, but actual dynamics are a time-varying system coupled with wind, waves, currents, and ship characteristics. Existing technologies ignore detailed models of environmental factors and coupling effects, leading to deviations between the model and the physical process. In complex operating conditions, prediction compensation is insufficient, tracking accuracy decreases, and the risk of signal interruption increases. 2. Passive response control caused by attitude prediction and compensation lag. The servo system calculates based on current or past errors, with lag compensation. Antenna adjustments "catch up" with pointing deviations. The lack of high-confidence multi-step prediction makes it impossible to pre-plan the optimal trajectory, leading to frequent sudden stops and starts, increased wear and energy consumption, and easy loss of tracking. 3. Fragmented control optimization dimensions leading to an inability to coordinate mechanical stability and communication performance. Optimization targets focus on minimizing errors or smoothing speed, neglecting to maintain signal strength. Geometric accuracy does not equate to optimal quality, as signals are affected by multipath propagation, obstruction, and sidelobes. The lack of a mechanism to incorporate signal strength distribution into the control loop may cause the antenna to pass through weak signal areas or miss strong signal points, creating a conflict between mechanical optimization and communication quality. Therefore, intelligent control methods integrating multi-dimensional environmental perception, digital twin simulation, data-driven prediction, and multi-objective collaborative optimization are needed to overcome these shortcomings and achieve a technological leap from passive response to proactive planning for optimal tracking. Summary of the Invention

[0004] The purpose of this invention is to provide a shipborne servo system control method and system based on multi-dimensional data analysis to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides a shipborne servo system control method based on multi-dimensional data analysis, comprising:

[0006] S100 collects historical records and 3D models of the ship, as well as monitoring data and environmental characteristics during navigation. A virtual simulation model is built based on the 3D model and dynamically mapped using real-time collected data.

[0007] Historical records include monitoring data and environmental characteristic data at different times.

[0008] A three-dimensional model is a digital twin that can be used for simulation analysis, constructed using computer modeling technology based on the geometric features and structural relationships of a ship's physical structure.

[0009] Monitoring data includes bearing data, signal data, and attitude data. Bearing data refers to the spatial orientation of the communication target relative to the ship. Signal data refers to the signal strength obtained by adjusting the antenna mount on the ship's attitude.

[0010] Attitude data includes hull attitude and antenna mount attitude. Hull attitude includes pitch, roll, and yaw. Antenna mount attitude includes azimuth and elevation.

[0011] Environmental characteristic data includes wind characteristic data, water characteristic data, and mobility data.

[0012] Wind characteristic data refers to a set of statistical parameters that characterize the spatiotemporal distribution patterns and turbulent characteristics of natural wind fields, covering the mean wind speed and direction, fluctuation spectrum, coherence structure, and their interaction effects with topography and atmospheric boundary layer.

[0013] Wind dynamics data represents external environmental loads that add extra environmental disturbances and loads to a ship's navigation, affecting course maintenance, stability, and directly contributing to drag.

[0014] Water characteristic data refers to the set of dynamic characteristic parameters describing the movement of water bodies such as waves, currents and tides, including wave spectrum energy distribution, vertical structure of flow field, and nonlinear wave interaction and wave-current interaction mechanism.

[0015] Hydrodynamic data is the underlying foundation, determining how a ship moves in waves and what forces it experiences. It forms the physical basis for analyzing all marine performance.

[0016] Motion data refers to the set of quantitative relationships between the conversion of energy into speed and maneuverability of a ship's propulsion system under different operating conditions, covering the drag-thrust coupling mechanism, propeller hydrodynamic characteristics, and energy efficiency mapping function.

[0017] Motion data is the core application, specifically focusing on the direction of movement during motion, and conducting in-depth research on how to optimize the hull and propulsion system to achieve the required speed with minimal energy consumption. Resistance is the most critical hydrodynamic component.

[0018] The specific steps for building, training, and dynamically mapping a virtual simulation model include:

[0019] S101. Based on the three-dimensional model of the ship, an independent hydrodynamic model, and an atmospheric wind field model, a virtual simulation model consisting of the ship model, water body model, and wind field model is built in the simulation environment.

[0020] Among them, the ship model is constructed based on geometric features and structural relationships, while the water body model and wind field model respectively represent the physical laws of water feature data and wind feature data.

[0021] S102. Based on the azimuth and attitude data in the real-time monitoring data, analyze the location and pointing direction of the antenna mount on the ship model, and set the communication target in the relative spatial orientation of the ship model in the virtual simulation model.

[0022] S103. Real-time acquisition of environmental characteristic data and monitoring data drives the dynamic updating of wind field model, water body model and ship model in virtual simulation model, and completes real-time mapping of water and air environment, ship attitude and antenna mount pointing in reality.

[0023] By driving the physical model with real-time collected environmental feature data and using monitoring data as state input, the virtual model can reproduce the complex coupled dynamic processes of the physical world in real time.

[0024] A high-fidelity digital twin simulation environment synchronized with real ships and navigation environments was constructed, providing a near-realistic virtual experimental field for subsequent analysis and prediction.

[0025] S200: A virtual simulation model is trained using historical records to predict ship attitude changes over a future timeframe. The pointing trajectory of the antenna mount is derived, a two-dimensional curved trajectory surface is constructed, and the signal strength distribution is analyzed. Specifically, this includes:

[0026] S201. Train the virtual simulation model using historical data to construct a prediction model for ship attitude. Generate a predicted ship attitude dataset based on the prediction model, and then deduce the pointing demonstration trajectory in both spatiotemporal dimensions. Specifically, this includes:

[0027] S2011. Input historical records into the virtual simulation model, drive the water body model and wind field model with environmental characteristic data at the same time, and use the ship attitude data at the corresponding time as the verification benchmark.

[0028] S2012. Through physical simulation and data analysis, optimize the coupling parameters in the virtual simulation model, and train and calibrate the quantitative mapping relationship between environmental characteristic data and the six-degree-of-freedom motion response of the ship.

[0029] S2013. An algorithm model based on long short-term memory artificial neural networks is adopted, and historical ship attitude time series data is used as input for training to construct a predictive model that can predict the future attitude of ships.

[0030] Long Short-Term Memory (LSTM) neural networks excel at processing time-series data and can capture the long-term dependencies and dynamic patterns of ship motion under the disturbances of complex environments such as wind, waves, and currents. They are more adaptable to nonlinear marine environments than traditional linear prediction methods.

[0031] S2014. During ship navigation, real-time environmental characteristic data and current ship attitude are continuously collected and input into the prediction model for simulation and extrapolation to generate future duration. This dataset contains predicted ship attitudes. It includes at least predicted values ​​for the ship's pitch, roll, and heading angles.

[0032] S2015. Based on the predicted ship attitude dataset and combined with the fixed installation geometric relationship between the antenna mount and the ship hull in the three-dimensional model, the antenna mount pointing deviation caused by the change in ship attitude is calculated through coordinate transformation.

[0033] S016. Analyze the spatial orientation of the communication target on the ship model, and deduce the future duration in reverse by combining the pointing deviation. The antenna mount's attitude is adjusted to compensate for the ship's motion, thereby generating a pointing demonstration trajectory in both spatiotemporal dimensions.

[0034] S202. Analyze the maximum adjustment range of the azimuth and elevation angles in the antenna mount attitude, and construct a two-dimensional curved trajectory surface centered on the communication target's azimuth, setting a reference point. Specifically, this includes:

[0035] S2021. Obtain the maximum azimuth adjustment range and the maximum elevation adjustment range allowed by the mechanical structure of the antenna mount, and define the limit motion boundary of the antenna line of sight in inertial space based on these two ranges.

[0036] S2022. Discrete sampling is performed within the maximum azimuth and elevation adjustment range. For each combination of azimuth and elevation values ​​obtained from the sampling, the coordinates of the corresponding antenna line-of-sight pointing azimuth in inertial space are calculated.

[0037] Discrete sampling follows three principles: comprehensiveness, regularity, and purposefulness. These principles refer to:

[0038] Comprehensiveness: This means that the sampling points should cover the entire defined area.

[0039] Regularity: This refers to sampling being performed according to a predetermined and reasonable rule. For example, sampling at equal or varying angular intervals is conducted within a two-dimensional plane formed by the maximum and minimum values ​​of azimuth and elevation angles to ensure that the distribution of points on the trajectory surface is representative.

[0040] Purpose: The purpose of sampling is to construct or represent a continuous two-dimensional surface using a finite, representative set of discrete points. This is the core of constructing a continuous model using discrete points.

[0041] S2023. All coordinate points are spatially assembled to form a two-dimensional curved trajectory surface that characterizes the reachable range of the antenna's line of sight. This two-dimensional curved trajectory surface completely covers all possible directions of the antenna under mechanical constraints.

[0042] S2024. Combine the midpoint between the maximum azimuth adjustment range and the maximum elevation adjustment range as the reference direction, and set the three-dimensional spatial coordinate point corresponding to the reference direction as the center point of the two-dimensional curved trajectory surface.

[0043] S2025. The center point serves as an equivalent substitute for the communication target's spatial orientation relative to the ship's location, extracting historical data. The antenna's line-of-sight direction is analyzed based on the monitoring data at each moment and the attitude data at the same moment.

[0044] S2026. Analyze spatial orientation based on orientation data, and mark the times when the pointing orientation and spatial orientation are the same. After arranging all times in chronological order, use the marked times as separators to divide the time zones sequentially according to the order of the times.

[0045] S2027, Filter out times with a greater than and less than The time zone has only one earliest marked time, and the signal strength at the marked time is taken as the best signal strength for the corresponding time zone.

[0046] S2028. Analyze the directional and spatial orientation at each time point, regard the center point as the coordinate point of the spatial orientation, and map the corresponding coordinate points of the directional orientation at each time point on the two-dimensional curved trajectory surface as reference points.

[0047] A two-dimensional curved trajectory surface refers to a two-dimensional curved surface formed by the movement of the tip of a shipborne antenna in space as the azimuth and elevation angles change. It is a complex dynamic surface determined by the kinematics of the antenna servo system and the disturbance of the carrier.

[0048] S203. Based on the historical data, fit the relationship between the center point and each reference point, and combine it with the current signal strength to analyze the signal strength distribution trend of the two-dimensional curved trajectory surface. Specifically, this includes:

[0049] S2031. Analyze each reference point on the two-dimensional curved trajectory surface at each time point, obtain the signal strength in the signal data at each time point as the actual signal strength, and the optimal signal strength in the time zone where the time point is located.

[0050] S2032, Statistical Reference Points Corresponding number of time points The actual signal strength at each moment is used as the dependent variable, and the best signal strength in the time zone at that moment is used as the independent variable. The dependent and independent variables at each moment are packaged into a sample.

[0051] S2033, this The sample is input into the regression model, and the reference point is obtained by fitting. The nonlinear relationship is obtained by fitting the nonlinear relationship to each reference point.

[0052] S2034. Obtain the signal strength from the latest signal data, substitute it into the nonlinear relationship of each reference point to calculate the expected signal strength, and map it onto the corresponding positions of the reference points on the two-dimensional curved trajectory surface to obtain the signal strength distribution.

[0053] The core of constructing this two-dimensional curved trajectory surface is to comprehensively cover the entire adjustable range of azimuth and elevation angles through discrete sampling. For each sampling point, its nonlinear relationship is fitted using historical data.

[0054] The two-dimensional curved trajectory surface is equivalent to an empirical data model, describing "the expected signal strength that can be obtained when the antenna is pointed at this position, under the combined influence of the current ship attitude, environmental disturbances, and target orientation." This enables subsequent antenna mounting pointing path planning to shift from passively responding to signal fluctuations to actively planning high-signal-quality paths.

[0055] It enables accurate prediction of the future motion of ships and, based on this, derives the initial tracking trajectory of the antenna to compensate for the ship's motion.

[0056] Meanwhile, by learning from historical data, a mapping model between any pointing point within the entire mechanical reach of the antenna and the expected signal strength is constructed, providing a crucial "signal quality map" for subsequent trajectory optimization.

[0057] S300: Analyze the angular velocity change curve based on the pointing demonstration trajectory and divide the abnormal area. Establish different adjustment areas based on each abnormal area and randomly assemble a scheme.

[0058] The stability index of each scheme is calculated and a control scheme is selected. Based on the signal strength distribution, the pointing demonstration trajectory for each control zone within the control scheme is replanned, and the trajectories are then stitched together to obtain a new pointing demonstration trajectory. Specifically, this includes:

[0059] S301. Analyze the antenna mount attitude at different times in the demonstration trajectory, thereby analyzing the change of the antenna's line-of-sight pointing azimuth coordinates in inertial space and calculating the rate of angle change at different times.

[0060] S302. Draw the angular velocity change curve based on the fluctuation of the angular change rate over continuous time. Mark the part of the angular velocity change curve where the angular change rate is greater than a preset threshold as abnormal, and take the time interval corresponding to the abnormal line as the abnormal area.

[0061] S303. By expanding the abnormal area, different time periods are established for each abnormal line at both ends. The average angle change rate of each time period in the direction of the demonstration trajectory is recalculated. The time period with an average angle change rate not greater than a preset threshold is used as the adjustment area.

[0062] By analyzing the angular velocity change curve and setting a threshold to identify abnormal areas, it is equivalent to identifying the risky periods when the servo mechanism needs to undergo drastic acceleration and deceleration, which may cause vibration or tracking instability.

[0063] Subsequently, an adjustment region is established by expanding the anomaly region, which essentially defines a time window for the optimization algorithm to be replanned.

[0064] S304. Count the number of all abnormal areas. and establish There are several plans, each containing... There are several adjustment zones originating from different anomaly lines. Within the same scheme, the corresponding time periods of all adjustment zones do not intersect, and the adjustment zones in different schemes are not entirely the same.

[0065] S305. Calculate the stability index for each scheme, and select the scheme with the highest stability index as the adjustment scheme. Replace the corresponding abnormal area with the adjustment area in each adjustment scheme, and map each adjustment area onto the time dimension pointing to the demonstration trajectory. Specifically, this includes:

[0066] S3051. Evenly distribute sampling points on the angular velocity change curve, analyze the angular change rate at each sampling point, and calculate the standard deviation of the angular change rate for all sampling points. .

[0067] S3052. By analyzing the antenna mount attitude change in each adjustment zone within the demonstration trajectory analysis scheme, and combining the corresponding duration of the adjustment zone, the average angle change rate is calculated.

[0068] S3053. Map the average angle change rate of all adjustment zones to the corresponding time period in the speed change curve and make adjustments. Set the collection points evenly and analyze the angle change rate of each collection point.

[0069] S3054. Calculate the standard deviation of the rate of angle change for all data acquisition points. Preset constants and weighting coefficients and Substitute the values ​​into the formula to calculate the stability index of each scheme. :

[0070] ;

[0071] In the formula, For the first The average rate of angle change in each adjustment zone For the first The average angular change rate of each adjustment zone corresponds to the abnormal zone.

[0072] In the stability index calculation formula, the logarithmic term is used to evaluate the degree of improvement of the overall angular velocity stability of the entire scheme. The summation term is used to evaluate the suppression effect on the average angular velocity within each adjustment zone.

[0073] Through weighting coefficients and The emphasis of the two items can be adjusted. Selecting the scheme with the highest stability index from multiple randomly generated schemes is an efficient heuristic search strategy.

[0074] S306. Analyze the duration of each adjustment zone and the antenna mount attitude changes in the pointing demonstration trajectory. Combined with the signal strength distribution, replan the pointing demonstration trajectory for each adjustment zone, and stitch them together to obtain a new pointing demonstration trajectory. Specifically, this includes:

[0075] S3061, Analysis and Adjustment Area time period and duration and in the time period pointing to the demonstration trajectory The initial and final antenna mount attitudes are planned. A demonstration path pointing to the spacetime dimension.

[0076] S3062, The time required for each path pointing to the demonstration is... Furthermore, the antenna mount attitude change satisfies the condition that, starting from the initial antenna mount attitude, over a duration of [duration missing]. The subsequent development of the final antenna mount attitude.

[0077] S3063. Using the spatial orientation of the communication target at different times in the pointing demonstration path as the center point, the pointing orientation path is mapped on a two-dimensional curved trajectory surface according to the coordinate point of the pointing orientation of the antenna line of sight in inertial space.

[0078] S3064. Calculate the average signal strength of all reference points traversed by the directional path, and use it as the efficiency coefficient of the corresponding directional demonstration path.

[0079] S3065. Select the demonstration path with the highest efficiency coefficient and replace the time period in the demonstration trajectory. The partial pointing demonstration trajectory is then replaced by the partial pointing demonstration trajectory for each adjustment zone during the corresponding time period, and the two trajectories are then spliced ​​together to obtain a new pointing demonstration trajectory.

[0080] Within the adjustment zone, the trajectory is not simply smoothed out, but rather multiple candidate paths from the starting point to the end point are planned based on the signal strength distribution.

[0081] The average signal strength of each path passing through reference points is calculated as the efficiency coefficient, and the highest value is selected. This enables the path to actively detour to a pointing area with better signal strength under given time and space constraints, upgrading trajectory optimization from guaranteed tracking to high-quality tracking.

[0082] Based on the initial compensation trajectory, intelligent and multi-objective optimization is performed to generate a new directional demonstration trajectory that can both smooth the servo mechanism's movements and actively select high signal strength areas, thereby achieving the optimal balance between mechanical stability and communication quality.

[0083] S400 controls the antenna mounts on the ship to adjust their attitude according to the new pointing demonstration trajectory, records monitoring data and environmental characteristic data during navigation in real time, and stores them in the historical record.

[0084] The optimized algorithm results are transformed into actual antenna servo control commands to complete high-precision and high-stability target tracking tasks. Through continuous data acquisition, a closed loop of "perception-prediction-optimization-control-re-perception" is formed, enabling the system to have continuous learning and adaptive capabilities.

[0085] The present invention also provides a shipborne servo system control system based on multi-dimensional data analysis, including a multi-dimensional data perception module, a servo predictive analysis module, a pointing demonstration planning module, and an attitude adjustment control module.

[0086] The multi-dimensional data perception module is used to collect historical data, 3D models, monitoring data, and environmental characteristic data of the ship. A virtual simulation model is then built and dynamically mapped.

[0087] By collecting ship historical records, 3D models, real-time monitoring data, and environmental characteristic data.

[0088] A virtual simulation model is built based on a 3D model and coupled with water and wind field models. Real-time data is used to drive the model to dynamically map ship attitude, antenna pointing and environmental conditions.

[0089] This enables digital twin modeling of ships and their navigation environment, providing a high-fidelity, multi-physics-based real-time simulation environment and data foundation for servo systems.

[0090] The servo predictive analysis module uses historical records to train a virtual simulation model, predicts the pointing trajectory of the antenna mount, constructs a two-dimensional curved trajectory surface, and analyzes the signal strength distribution.

[0091] A virtual simulation model was trained using historical records, and an LSTM neural network was used to construct a ship attitude prediction model, thereby deriving the future pointing demonstration trajectory of the antenna mount.

[0092] A two-dimensional curved trajectory surface is constructed by combining the maximum adjustment range of the antenna azimuth and elevation angles. The nonlinear relationship of each reference point is fitted by historical signal data, and the signal strength distribution trend is analyzed by combining the current signal strength.

[0093] Predicting the impact of ship motion on antenna pointing and pre-evaluating the communication signal quality under different pointing trajectories provides a basis for trajectory optimization.

[0094] The pointing demonstration planning module analyzes the angular velocity change curve based on the pointing demonstration trajectory and divides the abnormal area. It then establishes an adjustment zone for the abnormal area and sets up an adjustment scheme. Finally, it re-plans a new pointing demonstration trajectory based on the signal strength distribution.

[0095] The system analyzes the demonstration trajectory and plots the angular velocity change curve. After marking the abnormal area, it extends the time period to establish the adjustment zone and randomly constructs a scheme. The adjustment scheme is selected by calculating the stability index.

[0096] By combining the signal intensity distribution pattern on the two-dimensional curved trajectory surface, the optimal pointing demonstration path with the best efficiency coefficient is replanned for each adjustment zone, and spliced ​​together to form a new pointing demonstration trajectory.

[0097] It enables dynamic optimization of the antenna servo trajectory, maximizing signal stability while smoothing sudden changes in angular velocity and suppressing abnormal disturbances, thereby improving tracking accuracy and communication reliability.

[0098] The attitude adjustment control module controls the antenna mount on the ship to adjust its attitude according to the new pointing demonstration trajectory.

[0099] The antenna mounts on the control vessel are adjusted in real time according to the newly planned pointing demonstration trajectory, adjusting their azimuth and elevation angles. During the voyage, monitoring data and environmental characteristic data are continuously recorded and historical records are updated.

[0100] The system accurately executes optimized servo control commands to achieve stable target tracking by the antenna, and continuously accumulates learning samples through data closed-loop to support system self-adaptation and long-term performance optimization.

[0101] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0102] Comprehensive Perception and Modeling: The solution systematically integrates historical and real-time multidimensional data of the ship to construct a high-fidelity virtual simulation environment. It not only integrates the ship's own geometry and motion state but also deeply couples external flow fields and wind fields with environmental dynamics physical models, achieving real-time dynamic mapping of complex navigation situations. Breaking through the limitations of traditional methods that rely on single inertial feedback or simplified environmental assumptions, it provides a multiphysics digital twin environment that approximates the real world for basic analysis and decision-making.

[0103] Foresight in Prediction and Situational Awareness: The solution utilizes a trained time-series prediction model to quantitatively extrapolate the vehicle's motion attitude in future periods based on current and historical data, and then inversely derives the initial pointing trajectory required by the actuators to compensate for the vehicle's motion. By analyzing the reachable space of the actuators, a two-dimensional surface covering all possible pointing directions is constructed, and the distribution of signal intensity is mapped onto this surface based on historical signal data. This enables the system to possess dual forward-looking capabilities: predicting future motion and anticipating the communication performance at different pointing points.

[0104] Synergistic optimization of control: The core optimization mechanism of the solution lies in its departure from the isolated pursuit of a single indicator. First, by analyzing the kinematic curve of the initial trajectory, abnormal periods that may lead to violent actions of the actuator are identified. Then, by constructing and screening multiple solutions, the optimal balance is sought between smoothing sudden changes in angular velocity (improving mechanical stability) and actively selecting a pointing path with better signal strength (ensuring communication efficiency). This achieves a deep integration and synergistic optimization of the two major goals of optimizing mechanical performance and maximizing task efficiency. Attached Figure Description

[0105] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0106] Figure 1 This is a flowchart illustrating the shipborne servo system control method based on multi-dimensional data analysis according to the present invention.

[0107] Figure 2 This is a schematic diagram of the structure of the shipborne servo system control system based on multi-dimensional data analysis according to the present invention. Detailed Implementation

[0108] 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.

[0109] Example 1: Please refer to Figure 1 This invention provides a shipborne servo system control method based on multi-dimensional data analysis, including:

[0110] S100 collects historical records and 3D models of the ship, as well as monitoring data and environmental characteristics during navigation. A virtual simulation model is built based on the 3D model and dynamically mapped using real-time collected data.

[0111] Historical records include monitoring data and environmental characteristic data at different times.

[0112] In practice, the three-dimensional model is a digital twin constructed using computer modeling technology based on the geometric features and structural relationships of the ship's entity, and can be used for simulation analysis.

[0113] Monitoring data includes bearing data, signal data, and attitude data. Bearing data refers to the spatial orientation of the communication target relative to the ship. Signal data refers to the signal strength obtained by adjusting the antenna mount on the ship's attitude.

[0114] Attitude data includes hull attitude and antenna mount attitude. Hull attitude includes pitch, roll, and yaw. Antenna mount attitude includes azimuth and elevation.

[0115] In the specific implementation process, environmental characteristic data includes wind characteristic data, water characteristic data, and action force data.

[0116] Wind characteristic data refers to a set of statistical parameters that characterize the spatiotemporal distribution patterns and turbulent characteristics of natural wind fields, covering the mean wind speed and direction, fluctuation spectrum, coherence structure, and their interaction effects with topography and atmospheric boundary layer.

[0117] Wind dynamics data represents external environmental loads that add extra environmental disturbances and loads to a ship's navigation, affecting course maintenance, stability, and directly contributing to drag (air resistance).

[0118] Water characteristic data refers to the set of dynamic characteristic parameters describing the movement of water bodies such as waves, currents and tides, including wave spectrum energy distribution, vertical structure of flow field, and nonlinear wave interaction and wave-current interaction mechanism.

[0119] Hydrodynamic data is the underlying foundation, determining how a ship moves in waves and what forces (including resistance) it experiences. It is the physical basis for analyzing all marine performance characteristics (seakeeping, stability).

[0120] Motion data refers to the set of quantitative relationships between the conversion of energy into speed and maneuverability of a ship's propulsion system under different operating conditions, covering the drag-thrust coupling mechanism, propeller hydrodynamic characteristics, and energy efficiency mapping function.

[0121] Motion data is the core application, specifically focusing on the direction of movement during motion, and delving into how to optimize the hull and propulsion system to achieve the required speed with minimal energy consumption (fuel). Resistance is the most critical hydrodynamic component.

[0122] The specific steps for building, training, and dynamically mapping a virtual simulation model include:

[0123] S101. Based on the three-dimensional model of the ship, an independent hydrodynamic model, and an atmospheric wind field model, a virtual simulation model consisting of the ship model, water body model, and wind field model is built in the simulation environment.

[0124] Among them, the ship model is constructed based on geometric features and structural relationships, while the water body model and wind field model respectively represent the physical laws of water feature data and wind feature data.

[0125] S102. Based on the azimuth and attitude data in the real-time monitoring data, analyze the location and pointing direction of the antenna mount on the ship model, and set the communication target in the relative spatial orientation of the ship model in the virtual simulation model.

[0126] S103. Real-time acquisition of environmental characteristic data and monitoring data drives the dynamic updating of wind field model, water body model and ship model in virtual simulation model, and completes real-time mapping of water and air environment, ship attitude and antenna mount pointing in reality.

[0127] In the specific implementation process, by driving the physical model (wind field model, water body model) with real-time collected environmental feature data (such as wind spectrum and wave spectrum), and using the monitoring data (ship and antenna attitude) as state input, the virtual model can reproduce the complex coupled dynamic process of the physical world in real time.

[0128] A high-fidelity digital twin simulation environment synchronized with real ships and navigation environments was constructed, providing a near-realistic virtual experimental field for subsequent analysis and prediction.

[0129] S200: A virtual simulation model is trained using historical records to predict ship attitude changes over a future timeframe. The pointing trajectory of the antenna mount is derived, a two-dimensional curved trajectory surface is constructed, and the signal strength distribution is analyzed. Specifically, this includes:

[0130] S201. Train the virtual simulation model using historical data to construct a prediction model for ship attitude. Generate a predicted ship attitude dataset based on the prediction model, and then deduce the pointing demonstration trajectory in both spatiotemporal dimensions. Specifically, this includes:

[0131] S2011. Input historical records into the virtual simulation model, drive the water body model and wind field model with environmental characteristic data at the same time, and use the ship attitude data at the corresponding time as the verification benchmark.

[0132] S2012. Through physical simulation and data analysis, optimize the coupling parameters in the virtual simulation model, and train and calibrate the quantitative mapping relationship between environmental characteristic data and the six-degree-of-freedom motion response of the ship.

[0133] S2013. An algorithm model based on long short-term memory artificial neural networks is adopted, and historical ship attitude time series data is used as input for training to construct a predictive model that can predict the future attitude of ships.

[0134] Long Short-Term Memory (LSTM) neural networks excel at processing time-series data and can capture the long-term dependencies and dynamic patterns of ship motion under the disturbances of complex environments such as wind, waves, and currents. They are more adaptable to nonlinear marine environments than traditional linear prediction methods.

[0135] S2014. During ship navigation, real-time environmental characteristic data and current ship attitude are continuously collected and input into the prediction model for simulation and extrapolation to generate future duration. This dataset contains predicted ship attitudes. It includes at least predicted values ​​for the ship's pitch, roll, and heading angles.

[0136] S2015. Based on the predicted ship attitude dataset and combined with the fixed installation geometric relationship between the antenna mount and the ship hull in the three-dimensional model, the antenna mount pointing deviation caused by the change in ship attitude is calculated through coordinate transformation.

[0137] S016. Analyze the spatial orientation of the communication target on the ship model, and deduce the future duration in reverse by combining the pointing deviation. The antenna mount's attitude is adjusted to compensate for the ship's motion, thereby generating a pointing demonstration trajectory in both spatiotemporal dimensions.

[0138] S202. Analyze the maximum adjustment range of the azimuth and elevation angles in the antenna mount attitude, and construct a two-dimensional curved trajectory surface centered on the communication target's azimuth, setting a reference point. Specifically, this includes:

[0139] S2021. Obtain the maximum azimuth adjustment range and the maximum elevation adjustment range allowed by the mechanical structure of the antenna mount, and define the limit motion boundary of the antenna line of sight in inertial space based on these two ranges.

[0140] S2022. Discrete sampling is performed within the maximum azimuth and elevation adjustment range. For each combination of azimuth and elevation values ​​obtained from the sampling, the coordinates of the corresponding antenna line-of-sight pointing azimuth in inertial space are calculated.

[0141] In practice, discrete sampling follows three principles: comprehensiveness, regularity, and purposefulness. These principles refer to:

[0142] Comprehensiveness: This means that the sampling points should cover the entire defined range (i.e., the maximum adjustment range of azimuth and elevation angles).

[0143] Regularity: This refers to sampling being performed according to a predetermined and reasonable rule. For example, sampling at equal or varying angular intervals is conducted within a two-dimensional plane formed by the maximum and minimum values ​​of azimuth and elevation angles to ensure that the distribution of points on the trajectory surface is representative.

[0144] Purpose: The purpose of sampling is to construct or represent a continuous two-dimensional surface using a finite, representative set of discrete points. This is the core of constructing a continuous model using discrete points.

[0145] S2023. All coordinate points are spatially assembled to form a two-dimensional curved trajectory surface that characterizes the reachable range of the antenna's line of sight. This two-dimensional curved trajectory surface completely covers all possible directions of the antenna under mechanical constraints.

[0146] S2024. Combine the midpoint between the maximum azimuth adjustment range and the maximum elevation adjustment range as the reference direction, and set the three-dimensional spatial coordinate point corresponding to the reference direction as the center point of the two-dimensional curved trajectory surface.

[0147] S2025. The center point serves as an equivalent substitute for the communication target's spatial orientation relative to the ship's location, extracting historical data. The antenna's line-of-sight direction is analyzed based on the monitoring data at each moment and the attitude data at the same moment.

[0148] S2026. Analyze spatial orientation based on orientation data, and mark the times when the pointing orientation and spatial orientation are the same. After arranging all times in chronological order, use the marked times as separators to divide the time zones sequentially according to the order of the times.

[0149] S2027, Filter out times with a greater than and less than The time zone has only one earliest marked time, and the signal strength at the marked time is taken as the best signal strength for the corresponding time zone.

[0150] S2028. Analyze the directional and spatial orientation at each time point, regard the center point as the coordinate point of the spatial orientation, and map the corresponding coordinate points of the directional orientation at each time point on the two-dimensional curved trajectory surface as reference points.

[0151] In practical implementation, the two-dimensional curved trajectory surface refers to the two-dimensional curved surface formed by the movement of the tip of the shipborne antenna in space as the azimuth and elevation angles change. It is a complex dynamic surface determined by the kinematics of the antenna servo system and the carrier disturbance.

[0152] S203. Based on the historical data, fit the relationship between the center point and each reference point, and combine it with the current signal strength to analyze the signal strength distribution trend of the two-dimensional curved trajectory surface. Specifically, this includes:

[0153] S2031. Analyze each reference point on the two-dimensional curved trajectory surface at each time point, obtain the signal strength in the signal data at each time point as the actual signal strength, and the optimal signal strength in the time zone where the time point is located.

[0154] S2032, Statistical Reference Points Corresponding number of time points The actual signal strength at each moment is used as the dependent variable, and the best signal strength in the time zone at that moment is used as the independent variable. The dependent and independent variables at each moment are packaged into a sample.

[0155] S2033, this The sample is input into the regression model, and the reference point is obtained by fitting. The nonlinear relationship is obtained by fitting the nonlinear relationship to each reference point.

[0156] S2034. Obtain the signal strength from the latest signal data, substitute it into the nonlinear relationship of each reference point to calculate the expected signal strength, and map it onto the corresponding positions of the reference points on the two-dimensional curved trajectory surface to obtain the signal strength distribution.

[0157] The core of constructing this two-dimensional curved trajectory surface is to comprehensively cover the entire adjustable range of azimuth and elevation angles through discrete sampling. For each sampling point (reference point), its nonlinear relationship is fitted using historical data (e.g., through a regression model).

[0158] In practical implementation, the two-dimensional curved trajectory surface is equivalent to an empirical data model, describing "the expected signal strength that can be obtained when the antenna is pointed at this position, under the combined influence of the current ship attitude, environmental disturbances, and target orientation." This enables subsequent antenna mounting pointing path planning to shift from passively responding to signal fluctuations to actively planning high-signal-quality paths.

[0159] It enables accurate prediction of the future motion of ships and, based on this, derives the initial tracking trajectory of the antenna to compensate for the ship's motion.

[0160] Meanwhile, by learning from historical data, a mapping model (i.e., signal strength distribution pattern) between any pointing point within the entire mechanical reach of the antenna and the expected signal strength is constructed, providing a crucial "signal quality map" for subsequent trajectory optimization.

[0161] S300: Analyze the angular velocity change curve based on the pointing demonstration trajectory and divide the abnormal area. Establish different adjustment areas based on each abnormal area and randomly assemble a scheme.

[0162] The stability index of each scheme is calculated and a control scheme is selected. Based on the signal strength distribution, the pointing demonstration trajectory for each control zone within the control scheme is replanned, and the trajectories are then stitched together to obtain a new pointing demonstration trajectory. Specifically, this includes:

[0163] S301. Analyze the antenna mount attitude at different times in the demonstration trajectory, thereby analyzing the change of the antenna's line-of-sight pointing azimuth coordinates in inertial space and calculating the rate of angle change at different times.

[0164] S302. Draw the angular velocity change curve based on the fluctuation of the angular change rate over continuous time. Mark the part of the angular velocity change curve where the angular change rate is greater than a preset threshold as abnormal, and take the time interval corresponding to the abnormal line as the abnormal area.

[0165] S303. By expanding the abnormal area, different time periods are established for each abnormal line at both ends. The average angle change rate of each time period in the direction of the demonstration trajectory is recalculated. The time period with an average angle change rate not greater than a preset threshold is used as the adjustment area.

[0166] In the specific implementation process, abnormal areas are identified by analyzing the angular velocity change curve and setting a threshold. This is equivalent to identifying the risk period when the servo mechanism needs to undergo drastic acceleration and deceleration, which may cause vibration or tracking instability.

[0167] Subsequently, an adjustment region is established by expanding the anomaly region, which essentially defines a time window for the optimization algorithm to be replanned.

[0168] S304. Count the number of all abnormal areas. and establish There are several plans, each containing... There are several adjustment zones originating from different anomaly lines. Within the same scheme, the corresponding time periods of all adjustment zones do not intersect, and the adjustment zones in different schemes are not entirely the same.

[0169] S305. Calculate the stability index for each scheme, and select the scheme with the highest stability index as the adjustment scheme. Replace the corresponding abnormal area with the adjustment area in each adjustment scheme, and map each adjustment area onto the time dimension pointing to the demonstration trajectory. Specifically, this includes:

[0170] S3051. Evenly distribute sampling points on the angular velocity change curve, analyze the angular change rate at each sampling point, and calculate the standard deviation of the angular change rate for all sampling points. .

[0171] S3052. By analyzing the antenna mount attitude change in each adjustment zone within the demonstration trajectory analysis scheme, and combining the corresponding duration of the adjustment zone, the average angle change rate is calculated.

[0172] S3053. Map the average angle change rate of all adjustment zones to the corresponding time period in the speed change curve and make adjustments. Set the collection points evenly and analyze the angle change rate of each collection point.

[0173] S3054. Calculate the standard deviation of the rate of angle change for all data acquisition points. Preset constants and weighting coefficients and Substitute the values ​​into the formula to calculate the stability index of each scheme. :

[0174] ;

[0175] In the formula, For the first The average rate of angle change in each adjustment zone For the first The average angular change rate of each adjustment zone corresponds to the abnormal zone.

[0176] In the stability index calculation formula, the logarithmic term is used to evaluate the degree of improvement (reduction in standard deviation) of the overall angular velocity stability of the entire scheme. The summation term is used to evaluate the suppression effect on the average angular velocity within each adjustment zone.

[0177] Through weighting coefficients and The emphasis of the two items can be adjusted. Selecting the scheme with the highest stability index from multiple randomly generated schemes is an efficient heuristic search strategy.

[0178] S306. Analyze the duration of each adjustment zone and the antenna mount attitude changes in the pointing demonstration trajectory. Combined with the signal strength distribution, replan the pointing demonstration trajectory for each adjustment zone, and stitch them together to obtain a new pointing demonstration trajectory. Specifically, this includes:

[0179] S3061, Analysis and Adjustment Area time period and duration and in the time period pointing to the demonstration trajectory The initial and final antenna mount attitudes are planned. A demonstration path pointing to the spacetime dimension.

[0180] S3062, The time required for each path pointing to the demonstration is... Furthermore, the antenna mount attitude change satisfies the condition that, starting from the initial antenna mount attitude, over a duration of [duration missing]. The subsequent development of the final antenna mount attitude.

[0181] S3063. Using the spatial orientation of the communication target at different times in the pointing demonstration path as the center point, the pointing orientation path is mapped on a two-dimensional curved trajectory surface according to the coordinate point of the pointing orientation of the antenna line of sight in inertial space.

[0182] S3064. Calculate the average signal strength of all reference points traversed by the directional path, and use it as the efficiency coefficient of the corresponding directional demonstration path.

[0183] S3065. Select the demonstration path with the highest efficiency coefficient and replace the time period in the demonstration trajectory. The partial pointing demonstration trajectory is then replaced by the partial pointing demonstration trajectory for each adjustment zone during the corresponding time period, and the two trajectories are then spliced ​​together to obtain a new pointing demonstration trajectory.

[0184] In the actual implementation process, the trajectory is not simply smoothed in the adjustment area, but multiple candidate paths from the starting point to the end point are planned in combination with the signal strength distribution.

[0185] The average signal strength of each path passing through reference points is calculated as the efficiency coefficient, and the highest value is selected. This enables the path to actively detour to a pointing region with better signal strength under given time constraints (adjustment zone duration) and spatial constraints (start and end point attitudes), upgrading trajectory optimization from guaranteed tracking to high-quality tracking.

[0186] Based on the initial compensation trajectory, intelligent and multi-objective optimization is performed to generate a new directional demonstration trajectory that can both smooth the servo mechanism's movements (reduce sudden changes in angular velocity) and actively select high signal strength areas, thereby achieving the optimal balance between mechanical stability and communication quality.

[0187] S400 controls the antenna mounts on the ship to adjust their attitude according to the new pointing demonstration trajectory, records monitoring data and environmental characteristic data during navigation in real time, and stores them in the historical record.

[0188] In the specific implementation process, the optimized algorithm results are transformed into actual antenna servo control commands to complete the target tracking task with high precision and high stability. Through continuous data acquisition, a closed loop of "perception-prediction-optimization-control-re-perception" is formed, enabling the system to have continuous learning and adaptive capabilities.

[0189] Example 2: Please refer to Figure 2 The present invention also provides a shipborne servo system control system based on multi-dimensional data analysis, including a multi-dimensional data perception module, a servo predictive analysis module, a pointing demonstration planning module, and an attitude adjustment control module.

[0190] The multi-dimensional data perception module is used to collect historical data, 3D models, monitoring data, and environmental characteristic data of the ship. A virtual simulation model is then built and dynamically mapped.

[0191] In the specific implementation process, historical ship data, three-dimensional models, real-time monitoring data (position, signal, attitude) and environmental characteristic data (wind, water, mobility) are collected.

[0192] A virtual simulation model is built based on a 3D model and coupled with water and wind field models. Real-time data is used to drive the model to dynamically map ship attitude, antenna pointing and environmental conditions.

[0193] This enables digital twin modeling of ships and their navigation environment, providing a high-fidelity, multi-physics-based real-time simulation environment and data foundation for servo systems.

[0194] The servo predictive analysis module uses historical records to train a virtual simulation model, predicts the pointing trajectory of the antenna mount, constructs a two-dimensional curved trajectory surface, and analyzes the signal strength distribution.

[0195] In the specific implementation process, a virtual simulation model is trained using historical records and a ship attitude prediction model is constructed using an LSTM neural network, thereby deriving the future pointing demonstration trajectory of the antenna mount.

[0196] A two-dimensional curved trajectory surface is constructed by combining the maximum adjustment range of the antenna azimuth and elevation angles. The nonlinear relationship of each reference point is fitted by historical signal data, and the signal strength distribution trend is analyzed by combining the current signal strength.

[0197] Predicting the impact of ship motion on antenna pointing and pre-evaluating the communication signal quality under different pointing trajectories provides a basis for trajectory optimization.

[0198] The pointing demonstration planning module analyzes the angular velocity change curve based on the pointing demonstration trajectory and divides the abnormal area. It then establishes an adjustment zone for the abnormal area and sets up an adjustment scheme. Finally, it re-plans a new pointing demonstration trajectory based on the signal strength distribution.

[0199] In the specific implementation process, the pointing demonstration trajectory is analyzed and the angular velocity change curve is plotted. After marking the abnormal area, the adjustment area is established by extending the time period and randomly forming a scheme. The adjustment scheme is selected by calculating the stability index.

[0200] By combining the signal intensity distribution pattern on the two-dimensional curved trajectory surface, the optimal pointing demonstration path with the best efficiency coefficient is replanned for each adjustment zone, and spliced ​​together to form a new pointing demonstration trajectory.

[0201] It enables dynamic optimization of the antenna servo trajectory, maximizing signal stability while smoothing sudden changes in angular velocity and suppressing abnormal disturbances, thereby improving tracking accuracy and communication reliability.

[0202] The attitude adjustment control module controls the antenna mount on the ship to adjust its attitude according to the new pointing demonstration trajectory.

[0203] During the implementation process, the antenna mounts on the control vessel adjust their azimuth and elevation angles in real time according to the newly planned pointing demonstration trajectory, and continuously record monitoring data and environmental characteristic data and update historical records during navigation.

[0204] The system accurately executes optimized servo control commands to achieve stable target tracking by the antenna, and continuously accumulates learning samples through data closed-loop to support system self-adaptation and long-term performance optimization.

[0205] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0206] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A shipborne servo system control method based on multi-dimensional data analysis, characterized in that: The method includes: S100: Collect historical records and 3D models of the ship, as well as real-time monitoring data and environmental characteristic data collected during navigation; build a virtual simulation model based on the 3D model and perform dynamic mapping by combining the real-time collected data; S200: Train a virtual simulation model using historical records and predict the ship's attitude changes over a future period; derive the pointing demonstration trajectory of the antenna mount, construct a two-dimensional curved trajectory surface, and analyze the signal strength distribution. S300. Analyze the angular velocity change curve based on the pointing demonstration trajectory and divide it into abnormal regions. Establish different adjustment regions based on each abnormal region and randomly construct schemes. Calculate the stability index of each scheme and select an adjustment scheme. Replace the corresponding abnormal region with each adjustment region in the adjustment scheme and map each adjustment region onto the time dimension of the pointing demonstration trajectory. Combine the signal strength distribution trend to replan the pointing demonstration trajectory for each adjustment region in the adjustment scheme. After splicing, a new pointing demonstration trajectory is obtained. S400 controls the antenna mounts on the ship to adjust their attitude according to the new pointing demonstration trajectory, records monitoring data and environmental characteristic data during navigation in real time, and stores them in the historical record.

2. The shipborne servo system control method based on multi-dimensional data analysis according to claim 1, characterized in that: In S100, historical records include monitoring data and environmental characteristic data at different times; The three-dimensional model is a digital twin constructed using computer modeling technology based on the geometric features and structural relationships of the ship's entity, and can be used for simulation analysis. The monitoring data includes azimuth data, signal data, and attitude data; azimuth data refers to the spatial orientation of the communication target relative to the ship; signal data refers to the signal strength obtained by the antenna mount on the ship by adjusting its own attitude. Attitude data includes hull attitude and antenna mount attitude; hull attitude includes pitch, roll, and yaw angles; antenna mount attitude includes azimuth and elevation angles. Environmental characteristic data includes wind characteristic data, water characteristic data, and mobility data; Wind characteristic data refers to a set of statistical parameters that characterize the spatiotemporal distribution patterns and turbulent characteristics of natural wind fields, covering the mean wind speed and direction, fluctuation spectrum, coherence structure, and their interaction effects with topography and atmospheric boundary layer. Water characteristic data refers to the set of dynamic characteristic parameters describing the movement of water bodies such as waves, currents and tides, including wave spectrum energy distribution, vertical structure of the flow field, interaction between nonlinear waves and wave-current interaction mechanism; Action data refers to the set of quantitative relationships between the conversion of energy into speed and maneuverability of a ship's propulsion system under different operating conditions, covering the drag-thrust coupling mechanism, propeller hydrodynamic characteristics, and energy efficiency mapping function; The specific steps for building a virtual simulation model and dynamically mapping it include: S101. Based on the three-dimensional model of the ship, an independent hydrodynamic model and atmospheric wind field model, a virtual simulation model composed of the ship model, water body model and wind field model is built in the simulation environment. S102. Based on the azimuth and attitude data in the real-time monitoring data, analyze the location and pointing direction of the antenna mount on the ship model, and set the communication target in the relative spatial orientation of the ship model in the virtual simulation model. S103. Real-time acquisition of environmental characteristic data and monitoring data drives the dynamic updating of wind field model, water body model and ship model in virtual simulation model, and completes real-time mapping of water and air environment, ship attitude and antenna mount pointing in reality.

3. The shipborne servo system control method based on multi-dimensional data analysis according to claim 2, characterized in that: S200 includes: S201. Input historical records into the virtual simulation model for training and build a prediction model of ship attitude; generate a prediction ship attitude dataset based on the prediction model, and then reverse-engineer the pointing demonstration trajectory in the spatiotemporal dual dimensions. S202. Analyze the maximum adjustment range of the azimuth and elevation angles in the antenna mount attitude, construct a two-dimensional curved trajectory surface with the communication target azimuth as the center point, and set a reference point; S203. Based on the historical data, fit the relationship between the center point and each reference point, and combine it with the current signal strength to analyze the signal strength distribution of the two-dimensional curved trajectory surface.

4. The shipborne servo system control method based on multi-dimensional data analysis according to claim 3, characterized in that: S201 includes: S2011. Input historical records into the virtual simulation model, drive the water body model and wind field model with environmental characteristic data at the same time, and use the ship attitude data at the corresponding time as the verification benchmark. S2012. Through physical simulation and data analysis, optimize the coupling parameters in the virtual simulation model, and train and calibrate the quantitative mapping relationship between environmental characteristic data and the six-degree-of-freedom motion response of the ship. S2013. An algorithm model based on long short-term memory artificial neural network is adopted, and historical ship attitude time series data is used as input for training to build a prediction model that can predict the future attitude of ships. S2014. During ship navigation, real-time environmental characteristic data and current ship attitude are continuously collected and input into the prediction model for simulation and extrapolation to generate future duration. The dataset for predicting ship attitudes within the dataset; S2015. Based on the predicted ship attitude dataset and combined with the fixed installation geometric relationship between the antenna mount and the ship hull in the three-dimensional model, the antenna mount pointing deviation caused by the ship attitude change is calculated by coordinate transformation. S016. Analyze the spatial orientation of the communication target on the ship model, and deduce the future duration in reverse by combining the pointing deviation. The antenna mount's attitude is adjusted to compensate for the ship's motion, thereby generating a pointing demonstration trajectory in both spatiotemporal dimensions.

5. The shipborne servo system control method based on multi-dimensional data analysis according to claim 3, characterized in that: S202 includes: S2021. Obtain the maximum azimuth adjustment range and the maximum elevation adjustment range allowed by the mechanical structure of the antenna mount, and define the limit motion boundary of the antenna line of sight in inertial space based on these two ranges. S2022. Discrete sampling is performed within the maximum azimuth and elevation adjustment range. For each combination of azimuth and elevation values ​​obtained from the sampling, the coordinates of the corresponding antenna line-of-sight pointing azimuth in inertial space are calculated. S2023. All coordinate points are spatially assembled to form a two-dimensional curved trajectory surface that characterizes the reachable range of the antenna's line of sight. This two-dimensional curved trajectory surface completely covers all possible directions of the antenna under mechanical constraints. S2024. Combine the midpoint between the maximum azimuth adjustment range and the maximum elevation adjustment range as the reference direction, and set the three-dimensional spatial coordinate point corresponding to the reference direction as the center point of the two-dimensional curved trajectory surface. S2025. The center point serves as an equivalent substitute for the communication target's spatial orientation relative to the ship's location, extracting historical data. The antenna's line-of-sight direction is analyzed based on the monitoring data at each moment and the attitude data at the same moment. S2026. Analyze spatial orientation based on orientation data, and mark the times when the pointing orientation and spatial orientation are the same; after arranging all times in chronological order, use the marked times as the dividing positions to divide the time zones in the order of the times; S2027, Filter out times with a greater than and less than The time zone has only one earliest marked time, and the signal strength at the marked time is taken as the best signal strength in the corresponding time zone. S2028. Analyze the directional and spatial orientation at each time point, regard the center point as the coordinate point of the spatial orientation, and map the corresponding coordinate points of the directional orientation at each time point on the two-dimensional curved trajectory surface as reference points.

6. The shipborne servo system control method based on multi-dimensional data analysis according to claim 5, characterized in that: S203 includes: S2031. Analyze each reference point on the two-dimensional curved trajectory surface at each time point, obtain the signal strength in the signal data at each time point as the actual signal strength, and the optimal signal strength in the time zone where the time point is located. S2032, Statistical Reference Points Corresponding number of time points The actual signal strength at each moment is used as the dependent variable, and the best signal strength in the time zone at that moment is used as the independent variable. The dependent and independent variables at each moment are packaged into a sample. S2033, this The sample is input into the regression model, and the reference point is obtained by fitting. The nonlinear relationship is obtained by fitting nonlinear relationships to each reference point; and so on, the nonlinear relationships are obtained by fitting nonlinear relationships to each reference point respectively. S2034. Obtain the signal strength from the latest signal data, substitute it into the nonlinear relationship of each reference point to calculate the expected signal strength, and map it onto the corresponding positions of the reference points on the two-dimensional curved trajectory surface to obtain the signal strength distribution.

7. The shipborne servo system control method based on multi-dimensional data analysis according to claim 6, characterized in that: The S300 includes: S301. Analyze the antenna mount attitude at different times in the demonstration trajectory, thereby analyzing the change of the coordinate point of the antenna line of sight pointing in inertial space, and calculating the angle change rate at different times. S302. Draw the angular velocity change curve based on the fluctuation of the angular change rate over continuous time. Mark the part of the angular velocity change curve whose angular change rate is greater than a preset threshold as abnormal. The time interval corresponding to the abnormal line is taken as the abnormal area. S303. By expanding the abnormal area, different time periods are established for each abnormal line at both ends. The average angle change rate of each time period in the direction of the demonstration trajectory is recalculated. The time period with an average angle change rate not greater than the preset threshold is used as the adjustment area. S304. Count the number of all abnormal areas. and establish There are several plans, each containing... There are adjustment zones from different anomaly lines; within the same scheme, all adjustment zones do not intersect in the corresponding time periods, and all adjustment zones in different schemes are not completely the same; S305. Calculate the stability index of each scheme respectively, and take the scheme with the largest stability index as the adjustment scheme; replace the corresponding abnormal area with each adjustment area in the adjustment scheme, and map each adjustment area on the time dimension pointing to the demonstration trajectory. S306. Analyze the duration of each adjustment zone and the changes in the antenna mount attitude in the pointing demonstration trajectory. Combine the signal strength distribution trend to re-plan the pointing demonstration trajectory for each adjustment zone. After splicing, a new pointing demonstration trajectory is obtained.

8. The shipborne servo system control method based on multi-dimensional data analysis according to claim 7, characterized in that: S305 includes: S3051. Evenly distribute sampling points on the angular velocity change curve, analyze the angular change rate at each sampling point, and calculate the standard deviation of the angular change rate for all sampling points. ; S3052. By analyzing the antenna mount attitude change of each adjustment zone in the demonstration trajectory analysis scheme, and combining the corresponding duration of the adjustment zone, the average angle change rate is calculated. S3053. Map the average angle change rate of all adjustment zones to the corresponding time period in the speed change curve and make adjustments. Set the collection points evenly and analyze the angle change rate of each collection point. S3054. Calculate the standard deviation of the rate of angle change for all data acquisition points. Preset constants and weighting coefficients and Substitute the values ​​into the formula to calculate the stability index of each scheme. : ； In the formula, For the first The average rate of angle change in each adjustment zone For the first The average angular change rate of each adjustment zone corresponds to the abnormal zone.

9. The shipborne servo system control method based on multi-dimensional data analysis according to claim 7, characterized in that: S306 includes: S3061, Analysis and Adjustment Area time period and duration and in the time period pointing to the demonstration trajectory The initial and final antenna mount attitudes are planned. A demonstration path pointing to a point in both spacetime and time dimensions; S3062, The time required for each path pointing to the demonstration is... Furthermore, the antenna mount attitude change satisfies the condition that, starting from the initial antenna mount attitude, over a duration of [duration missing]. The subsequent development of the final antenna mount attitude; S3063. Using the spatial orientation of the communication target at different times in the pointing demonstration path as the center point, the pointing orientation path is mapped on a two-dimensional curved trajectory surface according to the coordinate point of the pointing orientation of the antenna line of sight in inertial space. S3064. Calculate the average signal strength of all reference points crossed by the directional path, and use it as the efficiency coefficient of the corresponding directional demonstration path; S3065. Select the demonstration path with the highest efficiency coefficient and replace the time period in the demonstration trajectory. The partial pointing demonstration trajectory is obtained by replacing the partial pointing demonstration trajectory of the corresponding time period in each adjustment zone and splicing them together to obtain a new pointing demonstration trajectory.

10. A shipborne servo system control system based on multi-dimensional data analysis, applied to the shipborne servo system control method based on multi-dimensional data analysis as described in claim 1, characterized in that: The system includes a multi-dimensional data perception module, a servo predictive analysis module, a pointing demonstration planning module, and an attitude adjustment control module; The multi-dimensional data perception module is used to collect historical records, 3D models, monitoring data, and environmental characteristic data of ships; and to build and dynamically map virtual simulation models. The servo predictive analysis module uses historical records to train a virtual simulation model, predicts the pointing trajectory of the antenna mount, constructs a two-dimensional curved trajectory surface, and analyzes the signal strength distribution. The pointing demonstration planning module analyzes the angular velocity change curve based on the pointing demonstration trajectory and divides the abnormal area. It then establishes an adjustment area for the abnormal area and sets up an adjustment scheme. Finally, it re-plans a new pointing demonstration trajectory based on the signal strength distribution. The attitude adjustment control module controls the antenna mount on the ship to adjust its attitude according to the new pointing demonstration trajectory.

Images