Unmanned ship marine target interception flight path shift changing strategy

Abstract

The invention discloses an unmanned ship marine target interception flight path shift change strategy. The strategy comprises the following steps: sensor data preprocessing: carrying out data format conversion and error correction on target information intercepted by a sensor of an unmanned ship; establishing a track prediction model: establishing the track prediction model to predict the future track of the unmanned ship based on the state information of the unmanned ship and the target and the marine environment parameters; planning a transition track: setting constraint conditions, and optimizing an algorithm to plan the transition track of the unmanned ship; and real-time adjustment and feedback: when the unmanned ship sails along the transition flight path, calculating the deviation between the actual flight path and the predicted flight path, adjusting the navigation parameters, and feeding back the adjustment information for re-predicting and planning the flight path. The unmanned ship marine target interception flight path shift change strategy aims at solving the problem of flight path abrupt change caused by unmanned ship perception information shift change, the flight path smoothness, the task adaptability and the system intelligence are improved, and the unmanned ship marine target interception task execution effect is enhanced.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned boat control, and in particular relates to a track handover strategy for unmanned boats to intercept targets at sea. Background Art

[0002] With the continuous development of ocean development and military applications, unmanned boats are increasingly used in maritime target interception missions. In the process of interception missions, the information acquisition and navigation control strategy of unmanned boats are crucial.

[0003] During the long-distance phase, the unmanned boat's sensors are limited by factors like range and accuracy, making it difficult to effectively acquire target information. However, the mother ship is typically equipped with more powerful detection equipment, such as high-performance radar and sonar, capable of accurately detecting key information such as the target's position, speed, and heading at long distances. Therefore, during this phase, the unmanned boat relies on the target information provided by the mother ship, navigating along a pre-planned, approximate path toward the target.

[0004] However, as the unmanned boat approaches its target and its own sensors (such as small radars and photoelectric sensors) are able to stably capture target information, it needs to transition from relying on information from the mother ship to relying on its own sensor information for navigation control, a process known as sensor information handover. This transition seems logical, but there are many problems in actual operation.

[0005] Firstly, the mother ship and the unmanned boat's own sensors differ in detection principles, accuracy, and update frequency. For example, the mother ship's radar may rely on high-power, long-wavelength detection, while the unmanned boat's optoelectronic sensors rely on optical imaging principles. The target position information obtained by each differs in data format and accuracy. This discrepancy can lead to significant deviations between the navigation instructions calculated by the unmanned boat based on newly acquired sensor information and the instructions previously executed based on information from the mother ship during the shift handover, resulting in a sudden change in the unmanned boat's trajectory.

[0006] Furthermore, the maritime environment is complex and ever-changing, with factors like wind, waves, and currents constantly impacting the navigation of unmanned vehicles. Even under normal navigation conditions, unmanned vehicles must constantly adjust their course and speed to maintain stability. The sudden changes in track caused by the handover of sensor information further exacerbate the instability of unmanned vehicles in complex sea conditions. In severe cases, the unmanned vehicle may deviate from its interception route, failing to complete the interception mission. It may even capsize due to drastic changes in direction or speed.

[0007] While some research has addressed the issue of track abrupt changes caused by the handover of sensor information during unmanned aerial vehicles (UAVs) intercepting targets at sea, most efforts focus on improving the performance of a single sensor or using simple data fusion algorithms. These efforts fail to systematically address the track abrupt changes caused by sensor differences and complex sea conditions during the handover process. For example, some studies simply perform a weighted average on sensor data. In the complex and ever-changing maritime environment, this approach cannot effectively address the differences in sensor characteristics and the interference of real-time sea conditions, making it difficult to ensure a smooth transition of the UAV's track. Furthermore, existing strategies often fail to fully consider the UAV's inherent dynamic characteristics and the actual navigation constraints at sea, leading to a disconnect between theory and practical application.

[0008] Therefore, there is an urgent need to develop a strategy that can effectively solve the problem of smooth track transition when unmanned boats intercept maritime target perception information. Summary of the Invention

[0009] Purpose of the invention: In order to overcome the above shortcomings, the purpose of the present invention is to provide an unmanned boat maritime target interception track handover strategy, which is reasonably designed and solves the problem of track mutation caused by the handover of perception information when the unmanned boat transitions from relying on the mother ship to provide target information to relying on its own sensor information to navigate, ensuring the smooth transition of the unmanned boat's track and improving the success rate of the unmanned boat's maritime target interception mission.

[0010] The purpose of the present invention is achieved through the following technical solutions:

[0011] A track handover strategy for intercepting targets at sea by an unmanned boat includes the following steps:

[0012] S1 sensor data preprocessing: converting the target information intercepted by the unmanned boat's own sensors into data format and performing error correction;

[0013] S2 track prediction model establishment: Based on the status information of the unmanned boat and the target and the marine environment parameters, a track prediction model is established to predict the future track of the unmanned boat;

[0014] S3 Transition Path Planning: Set constraints and optimize the algorithm to plan the transition path of the unmanned boat;

[0015] S4 real-time adjustment and feedback: When the unmanned boat is sailing along the transition track, the deviation between the actual track and the predicted track is calculated, the navigation parameters are adjusted, and the adjustment information is fed back to re-predict and plan the track.

[0016] Furthermore, the above-mentioned unmanned boat maritime target interception track handover strategy and the S1 sensor data preprocessing specifically include the following contents:

[0017] S11 Data Format Conversion: After the UAV's own sensor intercepts target information, the photoelectric sensor output image data is read in a specific format. The target features are extracted with the help of the image recognition library and converted into the target's position coordinate information in the UAV coordinate system. According to the radar sensor output data protocol, the received binary data is parsed into decimal physical quantities and converted into a data format in the UAV coordinate system that is unified with the photoelectric sensor. The binary data includes distance, speed, and angle.

[0018] S12 Error Correction: For photoelectric sensors, the extracted target position coordinates are corrected for distortion based on the lens distortion parameters. For radar sensors, the distance and speed data after analysis are compensated for errors using the ranging error model and the speed error model.

[0019] Preferably, the S11 data format conversion specifically includes the following: After the unmanned boat's own sensor intercepts the target information, the data of different types of sensors need to be converted into a unified format. If it is a photoelectric sensor, its output image data is read in a specific format (such as JPEG, PNG, etc.), and the target features are extracted with the help of an image recognition library (such as OpenCV) and converted into the position coordinate information of the target in the unmanned boat coordinate system. For radar sensors, according to their output data protocol (such as pulse Doppler radar protocol), the received binary data such as distance R, speed V, angle θ are parsed into decimal physical quantities and converted into a data format in the unmanned boat coordinate system that is unified with the photoelectric sensor.

[0020] Preferably, the S12 error correction specifically includes the following contents:

[0021] Photoelectric sensor: Based on the lens distortion parameters (obtained in the previous calibration), the extracted target position coordinates are corrected for distortion. Using the radial distortion correction formula, let the original coordinates be (x0, y0) and the corrected coordinates be (x, y), then x = x0 (1-k1r 2 +k2r 4 +k3r 6 ), y=y0(1+k1r 2 +k2r 4 +k3r 6 ),in k1, k2, and k3 are radial distortion coefficients, which are used to eliminate the influence of image deformation caused by the optical characteristics of the lens on the target position accuracy.

[0022] Radar sensor: Use the ranging error model (related to factors such as distance R and weather) and the speed error model (related to the radar signal processing algorithm) to compensate for the errors in the analyzed distance and speed data. Assume that the ranging error model is ΔR = a + bR + cW, where ΔR is the ranging error, a, b, c are model coefficients, and W is the weather influencing factor (such as rainfall, fog, and other quantifiable indicators). Based on the current measured distance R and real-time weather data, find the corresponding error correction value ΔR in the ranging error model and correct the distance data R to R. corrected =R-ΔR. The speed measurement error model is similar, assuming ΔV=d+eV+fs, where ΔV is the speed measurement error, d, e, f are model coefficients, and S is a parameter related to radar signal processing. Calculate ΔV based on actual conditions and correct the speed data V to V corrected =V-ΔV.

[0023] Furthermore, the above-mentioned unmanned boat maritime target interception track handover strategy and the establishment of the S2 track prediction model specifically include the following contents:

[0024] S21 parameter acquisition: real-time acquisition of the unmanned boat's navigation status information, including the unmanned boat's current position, speed, and heading, as well as pre-processed target motion status information, including position, speed, and acceleration, and current marine environment parameters;

[0025] S22 model construction: Based on the three-dimensional motion of the unmanned boat at sea, the motion equation of the unmanned boat is established according to Newton's second law, taking into account the external forces acting on the unmanned boat, including gravity, buoyancy, hydrodynamics, wind force, and ocean current force. Combined with the motion state of the target, the future track of the unmanned boat is predicted using the numerical integration method.

[0026] Preferably, the S21 parameter acquisition specifically includes the following contents:

[0027] Get the current position of the unmanned boat in real time (x uav ,y uav )(via GPS or other positioning systems), speed v uav (via Doppler speed meter or speed sensor), heading (via electronic compass or inertial navigation system) and other navigation status information, as well as the pre-processed target position (x target ,y target ), speed v target , acceleration a target At the same time, obtain the current marine environment parameters, such as wind speed v wind 、wind direction θ wind , ocean current speed v current and direction θcurrent etc. (can be obtained through meteorological sensors and current gauges).

[0028] Preferably, the S22 model is constructed as follows:

[0029] Based on the three-dimensional motion of the unmanned boat at sea, the motion equation of the unmanned boat is established according to Newton's second law. Considering that the unmanned boat is subject to gravity G = mg (m is the mass of the unmanned boat, g is the acceleration of gravity), buoyancy F b (According to Archimedes' principle F b =ρgV, ρ is the density of seawater, V is the volume of seawater displaced by the unmanned boat), hydrodynamic force (including resistance F drag , lift F lift etc.) and wind force F wind , ocean current force F current Other external forces.

[0030] In the horizontal direction, the motion equation of the unmanned boat can be expressed as:

[0031]

[0032] Among them, F x-thrust 、F y-thrust is the component of the thrust generated by the propulsion system in the x and y directions. Assume that the thrust of the propulsion system is F thrust , with the x-axis at an angle of α, then:

[0033] F x-thrust =F tharust cosα;

[0034] F y-thrust =F thrust sinα.

[0035] Among them, F x-drag 、F y-drag is the component of the hydrodynamic resistance in the x and y directions. Usually, the hydrodynamic resistance is proportional to the square of the velocity. Let the resistance coefficient be C d ,but:

[0036]

[0037] Among them, A is the projection area of ​​the unmanned boat in the direction of water flow, v uav，x 、v uav，y F is the speed component of the unmanned boat in the x and y directions; x-wind 、F y-wind is the component of wind force in the x and y directions, assuming that the wind force calculation formula is (ρ air is the air density, C wind is the wind force coefficient, A windis the projection area of ​​the unmanned boat in the wind direction), then:

[0038]

[0039] Among them, F x-current 、F y-current is the component of the ocean current force in the x and y directions, similar to the calculation of hydrodynamic resistance, assuming the ocean current force coefficient is C current ,but:

[0040]

[0041] Among them, v current，x 、v current，y are the x- and y-direction components of the ocean current velocity. By modeling and calculating these forces, combined with the target's motion state, numerical integration methods (such as the fourth-order Runge-Kutta method) are used to predict the future trajectory of the unmanned vehicle.

[0042] Furthermore, the above-mentioned unmanned boat maritime target interception track handover strategy, the S3 transition track planning, specifically includes the following contents:

[0043] S31 Constraint setting: setting the constraints related to track smoothness, navigation time, energy consumption and unmanned boat dynamics;

[0044] S32 optimization algorithm implementation: Using a genetic algorithm, the transition trajectory is represented as a set of coordinates of a series of intermediate points, an initial population is randomly generated, and the fitness value of each individual is calculated. The fitness function comprehensively considers the above constraints and uses genetic operations, including selection, crossover and mutation, to iteratively optimize the population and find the optimal transition trajectory.

[0045] Preferably, the S31 constraint condition setting specifically includes the following contents:

[0046] Track smoothness: The curvature change rate of the transition track is required to be within the acceptable range of the unmanned boat. Assume that the curvature of a point on the transition track is k and the curvature change rate is (s is the track arc length), limit ∈1 is the maximum curvature change rate that the unmanned boat can withstand, so as to avoid the impact of oversteering on the structure and navigation stability of the unmanned boat.

[0047] Navigation time: According to the mission requirements and target motion state, set a reasonable transition time upper limit T max Assume that the time it takes for the unmanned boat to travel from the starting point to the end point along the transition track is T, then T≤T max , ensuring that the unmanned boat can complete the handover in time and continue to approach the target.

[0048] Energy consumption: Considering the energy reserve and propulsion system efficiency of the unmanned boat, the energy consumption of the transition process is limited. Assume that the propulsion system power is P, the sailing time is T, the energy consumption E = PT, and the upper limit of energy consumption is E max , then E≤E max , ensuring that the unmanned boat has enough energy to complete subsequent interception missions.

[0049] Dynamic constraints: Consider the maximum steering angular velocity ω of the unmanned boat max , maximum acceleration a max and deceleration a min In transition trajectory planning, the steering angular velocity ω satisfies |ω|≤ω max , acceleration a satisfies a min ≤a≤a max .

[0050] Preferably, the implementation of the S32 optimization algorithm specifically includes the following contents:

[0051] Genetic algorithm is used to plan the transition trajectory. The transition trajectory is represented as a series of intermediate points (x i ,y i )(i=1, 2, ..., N) coordinates. An initial population is randomly generated, with each individual representing a possible transition trajectory. The fitness value of each individual is calculated. The fitness function, Fitness, comprehensively considers constraints such as trajectory smoothness, flight time, and energy consumption.

[0052] Track smoothness can be measured by calculating the change in curvature between adjacent intermediate points.

[0053] The sailing time can be calculated based on the speed of the unmanned boat v uav Distance from the middle point Calculated, assuming the navigation time is

[0054] Energy consumption can be calculated based on the propulsion system power model P = f(v uav ,ω)(f is the power function related to speed and steering angular velocity) and navigation distance estimation, assuming that the energy consumption is (P i is the corresponding segment power).

[0055] The fitness function can be expressed as Fitness = w1Smoothness + w2T + w3E (w1, w2, w3 are weight coefficients, which are adjusted according to actual task requirements).

[0056] Through genetic operations such as selection, crossover and mutation, the population is continuously iterated and optimized until the optimal individual that meets the constraints is found, that is, the optimal transition trajectory.

[0057] Furthermore, the above-mentioned unmanned boat maritime target interception track handover strategy, the S4 real-time adjustment and feedback, specifically includes the following contents:

[0058] S41 Deviation Calculation: When the UAV is navigating along the transition track, the actual position and heading information of the UAV is obtained at regular intervals, and compared with the position and heading at the corresponding moment of the predicted track, and the position deviation and heading deviation are calculated to determine the degree of deviation between the actual track and the predicted track;

[0059] S42 parameter adjustment: Based on the deviation size and direction, an adaptive control algorithm is used to calculate the heading angle increment and speed increment that need to be adjusted using the proportional-integral-derivative (PID) control algorithm, and send them to the propulsion system and steering system of the unmanned boat;

[0060] S43 feedback correction: The adjusted navigation status information is fed back to the track prediction model and the transition track planning module. The track prediction model re-predicts the future track, and the transition track planning module re-plans the subsequent transition track.

[0061] Preferably, the S41 deviation calculation specifically includes the following contents:

[0062] When the UAV is navigating along the transition track, the actual position (x actual ,y actual ) and heading Information, and the corresponding time position on the predicted track (x predicted ,y predicted ) and heading Compare. Position deviation is calculated using Euclidean distance The heading deviation is calculated as Determine the degree of deviation between the actual track of the unmanned boat and the predicted track.

[0063] Preferably, the S42 parameter adjustment specifically includes the following contents:

[0064] According to the deviation size and direction, the adaptive control algorithm is used to adjust the navigation parameters of the unmanned boat. Using the proportional-integral-differential (PID) control algorithm, the position deviation e position , heading deviation The PID controller output is:

[0065]

[0066] Among them, K p , K i , K d u is the proportional, integral and differential coefficient, which can be adjusted according to actual conditions. headingis the heading angle increment that needs to be adjusted, u speed The speed increment that needs to be adjusted is sent to the propulsion system and steering system of the unmanned boat to adjust its navigation status.

[0067] Preferably, the S42 parameter adjustment specifically includes the following contents:

[0068] The navigation status information after adjustment (such as the new position (x new ,y new ), speed v new ,course ) is fed back to the trajectory prediction model and the transition trajectory planning module. The trajectory prediction model re-predicts the future trajectory based on the new state information, and the transition trajectory planning module dynamically corrects the subsequent transition trajectory based on the new predicted trajectory and the current state to ensure that the unmanned boat always approaches the target along the optimal transition trajectory and adapts to the ever-changing maritime environment and target motion state. For example, in the trajectory prediction model, the new state is used as the initial condition, and numerical integration is re-calculated for the future trajectory; in the transition trajectory planning module, the new state is used as the starting point, the constraints and fitness function are reset, and the genetic algorithm is used to re-plan the subsequent transition trajectory.

[0069] Furthermore, in the above-mentioned unmanned boat maritime target interception track handover strategy, the data format conversion in S11 is to convert the radar sensor data into rectangular coordinates in the unmanned boat coordinate system using the polar coordinate to rectangular coordinate formula x=Rcosθ, y=Rsinθ.

[0070] For example, assuming that the radar outputs distance R and azimuth θ, they are converted into rectangular coordinates (x, y) in the UAV coordinate system using the polar coordinate to rectangular coordinate formula x = Rcosθ, y = Rsinθ.

[0071] Furthermore, the above-mentioned unmanned boat maritime target interception track handover strategy, the numerical integration method in S22 adopts the fourth-order Runge-Kutta method, for equation The iterative formula is:

[0072] k1=hf(t n , x n );

[0073]

[0074] k4=hf(t n +h,x n +k3);

[0075]

[0076] Where h is the time step, t n is the current time, xn The current state is used to calculate the future position of the unmanned boat through continuous iteration.

[0077] Furthermore, in the above-mentioned unmanned boat maritime target interception track handover strategy, the track smoothness in S3 is measured by calculating the curvature change between adjacent intermediate points, the navigation time is calculated based on the unmanned boat speed and the distance between the intermediate points, and the energy consumption is estimated based on the propulsion system power model and the navigation distance.

[0078] Specifically, let the adjacent middle points (x i ,y i ), (x i+1 ,y i+1 ), (x i+2 ,y i+1 ), according to the curvature formula (here The curvature k between two points can be calculated by approximating the difference of the values ​​of adjacent points i , then the curvature changes Δk i =|k i+1 -k i |, track smoothness index

[0079] Compared with the prior art, the present invention has the following beneficial effects:

[0080] (1) The track handover strategy for intercepting targets at sea by an unmanned boat disclosed in the present invention improves track smoothness. By preprocessing sensor data, data of different formats and precisions are converted into a unified format and errors are corrected, providing an accurate data basis for subsequent track planning. A track prediction model is established based on the motion state of the unmanned boat and the target, taking into account their dynamic characteristics and marine environmental factors, so that the predicted track is more in line with reality. With track smoothness as a constraint condition, a genetic algorithm is used to plan the transition track, avoiding track mutations caused by the handover of sensor information, ensuring that the unmanned boat track can smoothly transition, and thus improving its stability in navigation in complex marine environments.

[0081] (2) The unmanned boat maritime target interception track handover strategy disclosed in the present invention enhances mission adaptability; it comprehensively considers the dynamic characteristics of the unmanned boat, such as modeling various forces when establishing the motion equation, as well as marine environmental factors such as wind speed, wind direction, and ocean currents, making the strategy more suitable for actual application scenarios. Whether the target's motion state changes or encounters different complex sea conditions, the unmanned boat can better respond, thereby improving the success rate of the maritime target interception mission;

[0082] (3) The unmanned boat maritime target interception track handover strategy disclosed in this invention enhances the system's intelligence. It obtains the unmanned boat's navigation status information in real time and compares it with the predicted track. It uses an adaptive control algorithm (such as a PID control algorithm) to adjust navigation parameters in real time based on the deviation. The adjusted information is then fed back to the track prediction model and transition track planning module for dynamic correction. This series of operations enables adaptive control of the unmanned boat's navigation process, enhances the system's intelligence, reduces manual intervention, and improves mission execution efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is a flow chart of the unmanned boat maritime target interception track handover strategy described in the present invention. DETAILED DESCRIPTION

[0084] The following embodiments are combined with the attached Figure 1 The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts are within the scope of protection of the present invention.

[0085] The following embodiment provides a track handover strategy for intercepting targets at sea by an unmanned boat.

[0086] Example

[0087] like Figure 1 As shown, the unmanned boat maritime target interception track handover strategy of Example 1 includes the following contents:

[0088] (1) Sensor data preprocessing

[0089] 1. Data format conversion

[0090] When the UAV's own sensor intercepts the target information, the photoelectric sensor's output image data is usually stored in a specific image format (such as JPEG, PNG, etc.). The data parsing program is started and the image is processed using an image recognition library (such as OpenCV).

[0091] First, the image is loaded into memory, and then the target's features are extracted using an image recognition algorithm. For example, a deep learning-based object detection algorithm (such as YOLO or Faster R-CNN) can be used to identify the target's location in the image. Next, the target's pixel coordinates in the image are converted into the target's position in the UAV's coordinate system.

[0092] Radar sensors receive binary data such as distance R, velocity V, and angle θ, based on their output data protocols (e.g., pulse Doppler radar protocols). This binary data needs to be parsed into decimal physical quantities. For example, a radar data parsing program can convert the binary data into corresponding decimal values ​​based on the bit width and encoding method specified by the protocol. Then, using the polar coordinate to rectangular coordinate formula x = Rcosθ, y = Rsinθ, this data is converted into rectangular coordinates (x, y) in the same UAV coordinate system as the photoelectric sensor.

[0093] 2. Error correction

[0094] Photoelectric sensor: In the early stage, the lens of the photoelectric sensor needs to be calibrated to obtain the distortion parameters of the lens. After the target position coordinates are extracted, the radial distortion correction formula is used to correct the distortion. Let the original coordinates be (x0, y0) and the corrected coordinates be (x, y), then

[0095] x=x0(1+k1r 2 +k2r 4 +k3r 6 );

[0096] y=y0(1+k1r 2 +k2r 4 +k3r 6 ).

[0097] in, k1, k2, and k3 are the radial distortion coefficients. By calculating the value of r and substituting it into the formula, we can obtain the corrected coordinates, thereby eliminating the impact of image distortion caused by the optical characteristics of the lens on the target position accuracy.

[0098] Radar sensor: Establish ranging error model and speed error model. Assume that the ranging error model is ΔR = a + bR + cW. Among them, ΔR is the ranging error, a, b, c are model coefficients, and W is the weather influencing factor (such as rainfall, fog and other quantifiable indicators). According to the current measured distance R and real-time weather data, find the corresponding error correction value ΔR in the ranging error model and correct the distance data R to R corrected =R-ΔR.

[0099] The velocity error model is assumed to be ΔV = d + eV + fS, where ΔV is the velocity error, d, e, f are model coefficients, and S is a parameter related to radar signal processing. Based on the actual measured velocity V and related parameters S, ΔV is calculated and the velocity data V is corrected to V corrected =V-ΔV.

[0100] (2) Steps for establishing the track prediction model

[0101] 1. Parameter acquisition

[0102] Obtain the navigation status information of the unmanned boat in real time and obtain the current position of the unmanned boat through GPS or other positioning systems (x uav ,y uav ), obtain the speed v through Doppler speed meter or speed sensor uav , obtain heading through electronic compass or inertial navigation system

[0103] At the same time, the target position (x target ,y target ), speed v target , acceleration a target And other motion status information.

[0104] Use meteorological sensors and current meters to obtain current marine environmental parameters, such as wind speed v wind 、wind direction θ wind , ocean current speed v current and direction θ current wait.

[0105] 2. Model construction

[0106] Based on the three-dimensional motion of the unmanned boat at sea, the motion equation of the unmanned boat is established according to Newton's second law. The unmanned boat is subject to gravity G = mg (m is the mass of the unmanned boat, g is the acceleration of gravity), buoyancy F b (According to Archimedes' principle F b =ρgV, ρ is the density of seawater, V is the volume of seawater displaced by the unmanned boat), hydrodynamic force (including resistance F drag , lift F lift etc.) and wind force F wind , ocean current force F current Other external forces.

[0107] In the horizontal direction, the motion equation of the unmanned boat can be expressed as:

[0108]

[0109] Among them, the components of the thrust generated by the propulsion system in the x and y directions are: Assuming that the thrust of the propulsion system is F thrust , with the x-axis at an angle of α, then:

[0110] F x-thrust =F tharust cosα;

[0111] F y-thrust =F thrust sinα.

[0112] The components of the hydrodynamic resistance in the x and y directions: Usually the hydrodynamic resistance is proportional to the square of the velocity. Let the resistance coefficient be C d ,but:

[0113]

[0114] Among them, A is the projection area of ​​the unmanned boat in the direction of water flow, v uav，x 、v uav，y are the x and y components of the UAV velocity.

[0115] Components of wind force in the x and y directions: Assume that the wind force calculation formula is:

[0116]

[0117] Among them, ρ air is the air density, C wind is the wind force coefficient, A wind is the projection area of ​​the unmanned boat in the wind field direction, then:

[0118]

[0119] The components of the ocean current force in the x and y directions: Let the ocean current force coefficient be C current ,but:

[0120]

[0121] Among them, v current，x 、v current,y are the components of the ocean current velocity in the x and y directions.

[0122] By modeling and calculating these forces, combined with the motion state of the target, the future trajectory of the unmanned boat is predicted using numerical integration methods (such as the fourth-order Runge-Kutta method).

[0123] Taking the fourth-order Runge-Kutta method as an example, for the equation The iterative formula is: k1=hf(t n , x n );

[0124]

[0125] k4=hf(t n +h,x n +k3);

[0126]

[0127] Where h is the time step, t n is the current time, x n The current state is used to calculate the future position of the unmanned boat through continuous iteration.

[0128] (3) Transition trajectory planning steps

[0129] 1. Constraint setting

[0130] Track smoothness: Assume that the curvature of a point on the transition track is k, and the curvature change rate is (s is the track arc length), limit ∈1 is the maximum curvature change rate that the UUV can withstand. By discretizing the transition track and calculating the curvature change rate between adjacent discrete points, it can be ensured that it meets the constraint conditions to avoid oversteering affecting the UUV structure and navigation stability.

[0131] Navigation time: According to the mission requirements and target motion state, set a reasonable transition time upper limit T max Assume that the time it takes for the unmanned boat to travel from the starting point to the end point along the transition track is T, then T≤T max The navigation time of each section of the UAV on the transition track can be calculated and accumulated to obtain T, ensuring that it does not exceed the upper limit, so as to ensure that the UAV can complete the handover in time and continue to approach the target.

[0132] Energy consumption: Considering the energy reserve and propulsion system efficiency of the unmanned boat, the energy consumption of the transition process is limited. Assume that the propulsion system power is P, the sailing time is T, the energy consumption E = PT, and the upper limit of energy consumption is E max , then E≤E max The energy consumption can be estimated based on the power model of the propulsion system and the sailing distance to ensure that it does not exceed the upper limit and that the unmanned boat has enough energy to complete subsequent interception missions.

[0133] Dynamic constraints: Consider the maximum steering angular velocity ω of the unmanned boat max , maximum acceleration a max and deceleration a min In transition trajectory planning, the steering angular velocity ω satisfies |ω|≤ω max , acceleration a satisfies a min ≤a≤a max When planning the trajectory, the steering angular velocity and acceleration of each segment can be calculated and restricted to ensure that the dynamic constraints are met.

[0134] 2. Optimization algorithm implementation

[0135] Genetic algorithm is used to plan the transition trajectory. The transition trajectory is represented as a series of intermediate points (x i ,y i )(i=1, 2, ..., N) coordinate set.

[0136] The initial population is randomly generated, and each individual represents a possible transition trajectory.

[0137] Calculate the fitness value of each individual. The fitness function Fitness comprehensively considers constraints such as track smoothness, navigation time, and energy consumption.

[0138] Track smoothness: Set the adjacent middle points (x i ,y i ), (x i+1 ,y i+1 ), (x i+2 ,y i+2 ), according to the curvature formula (here The curvature k between two points can be calculated by approximating the difference of the values ​​of adjacent points i , then the curvature changes Δk i =|k i+1 -k i |, track smoothness index

[0139] Sailing time: can be adjusted according to the speed of the unmanned boat v uav Distance from the middle point Calculated, assuming the navigation time is

[0140] Energy consumption: can be calculated based on the propulsion system power model P = f(v uav ,ω)(f is the power function related to speed and steering angular velocity) and navigation distance estimation, assuming that the energy consumption is (P i is the corresponding segment power).

[0141] The fitness function can be expressed as Fitness = w1Smoothness + w2T + w3E (w1, w2, w3 are weight coefficients, which are adjusted according to actual task requirements).

[0142] Through genetic operations such as selection, crossover, and mutation, the population is continuously iterated and optimized until the optimal individual that satisfies the constraints is found, that is, the optimal transition trajectory. The selection operation can use the roulette wheel selection method, determining the probability of each individual being selected based on the size of the fitness value; the crossover operation can use single-point crossover or multi-point crossover to exchange genes of selected individuals; the mutation operation can randomly modify some of the genes of an individual to increase the diversity of the population.

[0143] (IV) Real-time adjustment and feedback steps

[0144] 1. Deviation calculation

[0145] When the UAV is navigating along the transition track, the actual position (x actual ,y actual ) and heading Information, and the corresponding time position on the predicted track (x predicted ,y predicted ) and heading contrast.

[0146] Position deviation is calculated using Euclidean distance The heading deviation is calculated as Determine the degree of deviation between the actual track of the unmanned boat and the predicted track.

[0147] 2. Parameter adjustment

[0148] According to the deviation size and direction, the adaptive control algorithm is used to adjust the navigation parameters of the unmanned boat. Using the proportional-integral-differential (PID) control algorithm, the position deviation e position =d position , heading deviation The PID controller output is:

[0149]

[0150] Among them, K p , K i , K d u is the proportional, integral and differential coefficient, which can be adjusted according to actual conditions. heading is the heading angle increment that needs to be adjusted, u speed The speed increment that needs to be adjusted is sent to the propulsion system and steering system of the unmanned boat to adjust its navigation status.

[0151] 3. Feedback and correction

[0152] The navigation status information after adjustment (such as the new position (x new ,y new ), speed v new ,course ) is fed back to the trajectory prediction model and transition trajectory planning module.

[0153] In the trajectory prediction model, the new state is used as the initial condition and the future trajectory is recalculated by numerical integration. For example, when using the fourth-order Runge-Kutta method, the new position and velocity are used as x n and the corresponding velocity components, and re-iterate the calculation.

[0154] In the transition trajectory planning module, the new state is used as the starting point, the constraints and fitness function are reset, and the subsequent transition trajectory is replanned using a genetic algorithm. This ensures that the unmanned boat always approaches the target along the optimal transition trajectory, adapting to the ever-changing marine environment and target motion state.

[0155] The unmanned boat maritime target interception track handover strategy described in Example 1 has the following innovative features:

[0156] 1. Systematically solve the problem of track mutation: Existing research has mostly focused on improving the performance of a single sensor or simple data fusion algorithms. The strategy described in this invention systematically solves the problem of track mutation caused by the handover of perception information of unmanned boats from four aspects: sensor data preprocessing, track prediction model establishment, transition track planning, and real-time adjustment and feedback, rather than just targeting a single link.

[0157] 2. Comprehensive consideration of multiple factors:

[0158] Sensor Characteristics: To address differences in detection principles, accuracy, and update frequency between the mother ship and the unmanned vehicle's sensors, specialized data analysis algorithms are employed to convert sensor data formats and perform error correction. For example, image recognition algorithms are used to extract target positions and correct distortion in photoelectric sensor image data, while radar sensor data is subjected to noise filtering and accuracy compensation based on its own principles.

[0159] Dynamic characteristics of unmanned vehicles: When establishing the trajectory prediction model, based on the current navigation state of the unmanned vehicle (position, speed, heading, etc.) and the target motion state, the unmanned vehicle's inertia, steering ability, acceleration and deceleration performance and other dynamic characteristics are considered. Newton's second law is combined with the principles of fluid mechanics to establish the motion equation of the unmanned vehicle in three-dimensional space:

[0160]

[0161] Among them, m is the mass of the unmanned boat, x and y are the position coordinates of the unmanned boat on the horizontal plane, and F x-thrust etc. are the force components in each direction, and track prediction is performed.

[0162] Marine environmental factors: When establishing the trajectory prediction model and transition trajectory planning, the effects of marine environmental factors such as wind, waves, and currents on the unmanned vehicle are fully considered. For example, when constructing the trajectory prediction model, environmental parameters such as wind speed, wind direction, and current speed and direction are incorporated into the unmanned vehicle's motion equations. When planning the transition trajectory, reasonable constraints are set based on these environmental factors.

[0163] 3. Use multiple optimization algorithms:

[0164] Track prediction: The track prediction model is established by combining nonlinear dynamics model with numerical integration method (such as fourth-order Runge-Kutta method) to predict the future track of the unmanned boat. The iterative formula is:

[0165] k1=hf(t n ,x n );

[0166]

[0167]

[0168] k4=hf(t n +h,x n +k3);

[0169]

[0170] Transition trajectory planning: Transition trajectory planning uses a genetic algorithm to search for the optimal transition trajectory, subject to constraints such as trajectory smoothness, flight time, and energy consumption. Track smoothness is measured by calculating the change in curvature between adjacent midpoints. Flight time is calculated based on the speed of the UAV and the distance between midpoints. Energy consumption is estimated based on the propulsion system power model and the flight distance.

[0171] 4. Real-time adaptive control: During the navigation process of the unmanned boat, the boat obtains its own sensor information and navigation status information in real time, and compares it with the predicted track. The boat uses the adaptive control algorithm (such as PID control algorithm) to adjust the navigation parameters (heading, speed) in real time according to the deviation, and feeds the adjusted information back to the track prediction model and transition track planning module for dynamic correction. The position deviation is calculated using the Euclidean distance. The heading deviation is calculated as To adapt to the ever-changing maritime environment and target motion status, and improve the system's intelligence level.

[0172] The present invention has many specific application paths, and the above is only a preferred embodiment of the present invention. It should be noted that the above embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention. For those skilled in the art, several improvements can be made without departing from the principles of the present invention, and these improvements should also be considered as the scope of protection of the present invention.

Claims

1. A track handover strategy for intercepting targets at sea by an unmanned boat, characterized in that: The steps include: S1 sensor data preprocessing: converting the target information intercepted by the UAV’s own sensors into data format and performing error correction; S2 track prediction model establishment: Based on the status information of the unmanned boat and the target and the marine environment parameters, a track prediction model is established to predict the future track of the unmanned boat; S3 Transition Path Planning: Set constraints and optimize the algorithm to plan the transition path of the unmanned boat; S4 real-time adjustment and feedback: When the unmanned boat is sailing along the transition track, the deviation between the actual track and the predicted track is calculated, the navigation parameters are adjusted, and the adjustment information is fed back to re-predict and plan the track.

2. The unmanned boat maritime target interception track handover strategy according to claim 1 is characterized in that: The S1 sensor data preprocessing specifically includes the following: S11 Data Format Conversion: After the UAV's own sensor intercepts target information, the photoelectric sensor output image data is read in a specific format. The target features are extracted with the help of the image recognition library and converted into the target's position coordinate information in the UAV coordinate system. According to the radar sensor output data protocol, the received binary data is parsed into decimal physical quantities and converted into a data format in the UAV coordinate system that is unified with the photoelectric sensor. The binary data includes distance, speed, and angle. S12 Error Correction: For photoelectric sensors, the extracted target position coordinates are corrected for distortion based on the lens distortion parameters. For radar sensors, the distance and speed data after analysis are compensated for errors using the ranging error model and the speed error model.

3. The unmanned boat maritime target interception track handover strategy according to claim 1 is characterized in that: The S2 track prediction model is established, specifically including the following contents: S21 parameter acquisition: real-time acquisition of the unmanned boat's navigation status information, including the unmanned boat's current position, speed, and heading, as well as pre-processed target motion status information, including position, speed, and acceleration, and current marine environment parameters; S22 model construction: Based on the three-dimensional motion of the unmanned boat at sea, the motion equation of the unmanned boat is established according to Newton's second law, taking into account the external forces acting on the unmanned boat, including gravity, buoyancy, hydrodynamics, wind force, and ocean current force. Combined with the motion state of the target, the future track of the unmanned boat is predicted using the numerical integration method.

4. The unmanned boat maritime target interception track handover strategy according to claim 1 is characterized in that: The S3 transition trajectory planning specifically includes the following: S31 Constraint setting: setting the constraints related to track smoothness, navigation time, energy consumption and unmanned boat dynamics; S32 optimization algorithm implementation: Using a genetic algorithm, the transition trajectory is represented as a set of coordinates of a series of intermediate points, an initial population is randomly generated, and the fitness value of each individual is calculated. The fitness function comprehensively considers the above constraints and uses genetic operations, including selection, crossover and mutation, to iteratively optimize the population and find the optimal transition trajectory.

5. The unmanned boat maritime target interception track handover strategy according to claim 1 is characterized in that: The S4 real-time adjustment and feedback specifically includes the following: S41 Deviation Calculation: When the UAV is navigating along the transition track, the actual position and heading information of the UAV is obtained at regular intervals, and compared with the position and heading at the corresponding moment of the predicted track, and the position deviation and heading deviation are calculated to determine the degree of deviation between the actual track and the predicted track; S42 parameter adjustment: Based on the deviation size and direction, an adaptive control algorithm is used to calculate the heading angle increment and speed increment that need to be adjusted using the proportional-integral-derivative (PID) control algorithm, and send them to the propulsion system and steering system of the unmanned boat; S43 feedback correction: The adjusted navigation status information is fed back to the track prediction model and the transition track planning module. The track prediction model re-predicts the future track, and the transition track planning module re-plans the subsequent transition track.

6. The unmanned boat maritime target interception track handover strategy according to claim 2 is characterized in that: The data format conversion in S11 is to use the polar coordinate to rectangular coordinate formula , , convert the radar sensor data into rectangular coordinates in the unmanned boat coordinate system.

7. The unmanned boat maritime target interception track handover strategy according to claim 3 is characterized in that: The numerical integration method in S22 adopts the fourth-order Runge-Kutta method. , and its iteration formula is: ; ; ; ; .

8. The unmanned boat maritime target interception track handover strategy according to claim 4 is characterized in that: The track smoothness in S3 is measured by calculating the curvature change between adjacent intermediate points. The navigation time is calculated based on the speed of the unmanned boat and the distance between the intermediate points. The energy consumption is estimated based on the propulsion system power model and the navigation distance.

9. The unmanned boat maritime target interception track handover strategy according to claim 5 is characterized in that: The position deviation in S41 is calculated using the Euclidean distance , the heading deviation is calculated as .

Images