Self-adaptive wind feedforward unmanned ship course control system based on active disturbance rejection

Abstract

The invention provides a self-adaptive wind feedforward unmanned ship course control system based on active disturbance rejection, and belongs to the technical field of unmanned ship course control. The system comprises an ultrasonic anemometer for collecting relative wind speed and relative wind direction; the gradient descent module outputs compensation gain based on the difference value between the actual course angle and the expected course angle; the wind feedforward module receives the relative wind speed, the relative wind direction and the compensation gain and outputs wind disturbance compensation; the expansion state observer outputs an actual course angle pre-estimated value, an actual course angle differential pre-estimated value and other unknown disturbance pre-estimated values; the non-linear state error feedback unit performs non-linear combination by using the state error to obtain an expected control quantity; the disturbance compensation unit determines the control quantity of the unmanned ship based on the expected control quantity, the wind disturbance compensation and other unknown disturbance estimation values. Rapid compensation of wind disturbance is achieved through self-adaptive wind feedforward control, and the response speed and stability of sailing direction control of the unmanned ship in a complex wind field are stably improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned vessel heading control technology, and in particular to an adaptive wind-feedback unmanned vessel heading control system based on active disturbance rejection. Background Technology

[0002] Unmanned surface vehicles (USVs), as an important carrier of modern marine technology, play a crucial role in marine resource exploration, environmental monitoring, and search and rescue. Their navigation performance hinges on course control capabilities, which are not only related to navigation safety but also fundamental to the execution of complex missions. Due to the dynamic and uncertain nature of the marine environment, USVs are affected by various disturbances such as wind, waves, and ocean currents during navigation, with wind disturbances having the most significant impact on hull motion. Therefore, researching high-precision course control methods is of great importance for improving the environmental adaptability and mission execution capabilities of USVs.

[0003] Currently, the heading control of unmanned surface vessels (USVs) mainly employs methods such as PID control, neural network control, and fuzzy control. However, these methods exhibit significantly reduced control performance when faced with the strongly coupled and nonlinear dynamic characteristics of USVs, especially in complex wind conditions. Active disturbance rejection control (ADRC), due to its independence from precise models and its disturbance estimation and compensation capabilities, has been widely used in heading control. However, in practical applications, when wind disturbances change drastically, the disturbance prediction of traditional ADRCs fluctuates, and the feedback loop's disturbance compensation is not timely, resulting in poor compensation effectiveness. Furthermore, USVs are generally lightweight and have a large wind-receiving area, leading to reduced rudder effectiveness at low speeds, further exacerbating the negative impact of wind disturbances on heading control, ultimately manifesting as increased track tracking deviation and decreased control stability. Summary of the Invention

[0004] In view of this, the present invention provides an adaptive wind-feedback unmanned surface vessel heading control system based on active disturbance rejection. Adaptive wind-feedback control is introduced into the active disturbance rejection control framework to achieve rapid compensation for wind disturbances. The gradient descent algorithm is used to adaptively adjust the gain of wind-feedback compensation to ensure the accuracy of compensation. The combination of the two effectively improves the response speed and accuracy of heading control of unmanned surface vessels in complex wind fields.

[0005] Therefore, the present invention provides the following technical solution: An adaptive wind-feedback unmanned surface vessel heading control system based on active disturbance rejection includes: Ultrasonic anemometer, wind feedforward module, gradient descent module, and active disturbance rejection control module; The active disturbance rejection control module includes: a tracking differentiator, an extended state observer, a nonlinear state error feedback unit, and a disturbance compensation unit; The ultrasonic anemometer collects relative wind speed and relative wind direction; The gradient descent module outputs a compensation gain based on the difference between the actual heading angle and the desired heading angle; The wind feedforward module receives relative wind speed, relative wind direction, and compensation gain to output wind disturbance compensation. The extended state observer outputs the estimated value of the actual heading angle, the estimated value of the derivative of the actual heading angle, and the estimated value of other unknown disturbances; The nonlinear state error feedback unit uses the state error to perform a nonlinear combination to obtain the desired control quantity; The disturbance compensation unit determines the unmanned vessel control quantity based on the desired control quantity, wind disturbance compensation, and other unknown disturbance estimates.

[0006] Furthermore, the gradient descent module outputs a compensation gain based on the difference between the actual heading angle and the desired heading angle, including:

[0007] In the formula, Represents learning; Represents the time step; Represents the gain at the current moment. Represents the gain from the previous moment; , This is the difference between the actual heading angle and the desired heading angle. This is the actual heading angle. The desired heading angle; Relative wind speed, relative wind direction; Represents air density; Represents the side projection area of ​​the hull above the waterline; Represents the length of the ship.

[0008] Furthermore, the wind feedforward module receives relative wind speed, relative wind direction, and compensation gain to output wind disturbance compensation, including:

[0009] in, Relative wind speed, relative wind direction; Represents air density; Represents the side projection area of ​​the hull above the waterline; Represents the length of the ship; To compensate for the gain; For wind disturbance compensation.

[0010] Furthermore, the extended state observer outputs the actual heading angle prediction, the actual heading angle derivative prediction, and other unknown disturbance predictions, including:

[0011] This represents a continuous power function with a linear segment near the origin; , represent The degree of nonlinearity of the function, ; represent Width of the linear interval of the function The function is represented as:

[0012] In the formula, The observation error represents the actual heading angle and the estimated actual heading angle; This represents the estimated actual heading angle; The estimated value representing the differential of the actual heading angle; This indicates the estimated value of other unknown disturbances; , , This represents the gain of the extended state observer.

[0013] Furthermore, the nonlinear state error feedback unit utilizes the state error to perform nonlinear combination to obtain the desired control quantity, including:

[0014] In the formula, Represents the proportionality coefficient. Represents the differential coefficient; , and This represents an adjustable parameter; To smooth out the deviation between the expected heading angle and the actual heading angle estimate; To smooth out the deviation between the estimated value of the derivative of the expected heading angle and the estimated value of the derivative of the actual heading angle after smoothing; This represents the expected control quantity.

[0015] Further, the tracking differentiator calculates the smoothed desired heading angle and the derivative of the smoothed desired heading angle, including:

[0016] The fhan function, representing the fastest control synthesis function, is defined as follows:

[0017] In the formula, the desired heading angle is expressed as: , This represents the expected heading angle after smoothing. This represents the derivative of the expected heading angle after smoothing. Represents the velocity factor. This represents the filter factor.

[0018] Furthermore, the disturbance compensation unit determines the unmanned vessel control quantity based on the desired control quantity, wind disturbance compensation, and other unknown disturbance estimates, including:

[0019] In the formula, Represents the equivalent input gain; Represents the amount of control over unmanned vessels; Represents the expected control quantity; This indicates the estimated value of other unknown disturbances; For wind disturbance compensation.

[0020] Advantages and positive effects of the present invention: This system ensures the effectiveness and stability of navigation control by tracking the set desired heading angle. The extended state observer predicts other unknown disturbances besides wind. The gradient descent module adaptively adjusts the compensation gain of the wind feedforward module, dynamically adjusting the feedforward compensation value. The actual input control quantity of the unmanned vessel is obtained by subtracting the wind feedforward compensation value and the estimated values ​​of other unknown disturbances in the disturbance compensation module.

[0021] 1) Adaptive wind feedforward control is introduced into the traditional active disturbance rejection control framework to construct a feedforward compensation channel that directly acts on the control torque. This forms a direct compensation mechanism for strong disturbance signals, avoiding the problem of untimely disturbance compensation caused by the lag in the response time of traditional active disturbance rejection feedback, and realizing rapid compensation for wind disturbances. This effectively improves the response speed and stability of heading control in complex wind fields.

[0022] 2) The gradient descent algorithm is used to adaptively adjust the compensation gain of the wind feedforward controller. A loss function based on the heading angle error is designed, and the momentum method is used to accelerate gradient convergence. This ensures that the feedforward gain can be quickly adjusted when the wind speed changes suddenly or fluctuates continuously, so as to achieve a better wind disturbance compensation effect and break through the limitations of traditional fixed gain feedforward compensation. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a block diagram of the heading control system of an adaptive wind-feedback unmanned vessel based on active disturbance rejection. Figure 2 This is a schematic diagram of the hardware-in-the-loop simulation platform structure in the embodiment; Figure 3 This is a comparison chart of the heading angle tracking performance of the unmanned vessel in the embodiment. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] Combination Figure 1 As shown, the present invention provides an adaptive wind-feedback unmanned surface vessel heading control system based on active disturbance rejection, comprising: an ultrasonic anemometer, a wind feedforward module, a gradient descent module, and an active disturbance rejection control module; the active disturbance rejection control module comprises: a tracking differentiator, an extended state observer, a nonlinear state error feedback unit, and a disturbance compensation unit.

[0028] 1. The wind feedforward module has three input signals and one output signal. The input signals include: relative wind speed and relative wind direction collected by the ultrasonic anemometer, and the compensation gain output by the gradient descent module. The output is the wind disturbance compensation, and the specific formula is as follows:

[0029] in, Relative wind speed, relative wind direction; Represents air density; Represents the side projection area of ​​the hull above the waterline; Represents the length of the ship; To compensate for the gain; For wind disturbance compensation.

[0030] 2. Gradient descent module, the input signal is the actual heading angle. and expected heading angle The difference between , is represented as: The output signal is the compensation gain of the wind feedforward module. The calculation process is as follows: Assumption 1: Other disturbances can be completely canceled out by the feedback component, and the heading angle error is mainly dominated by the wind disturbance feedforward term. Assumption 2: Ignoring the coupling between the differential reference term and the residual feedback term, the dominant part of the error dynamics is: ; .

[0031]

[0032] In the formula, The loss function represents the relationship between the compensation gain and the heading angle error. This represents the direction of gradient descent; The derivative representing the sensitivity of the system shows... How does it change with the gain k; b represents a constant coefficient; This represents the actual wind disturbance encountered in the direction of bow roll; The integral effect representing wind feedforward compensation; This represents the integral effect of the actual wind disturbance in the bow direction.

[0033] Focusing only on instantaneous error changes, it is assumed that: Finally, we obtain the complete gradient descent formula:

[0034] The momentum method is used to accelerate gradient convergence and reduce oscillations in local optima. The momentum gradient formula is expressed as:

[0035] In the formula, Represents the momentum coefficient; Represents the momentum gradient at the current moment; This represents the momentum gradient at the previous moment.

[0036] Using the derived gradient descent formula to compensate for gain The algorithm performs real-time updates and incorporates the constant b into the learning rate. The update formula is shown below:

[0037] In the formula, This represents the learning rate, which determines the size of each step the algorithm takes to reach the optimal solution. Represents the time step; Represents the compensation gain at the current moment. This represents the compensation gain from the previous moment.

[0038] 3. Tracking differentiator, the input signal is the desired heading angle given by the system. The output is the smoothed desired heading angle and its derivative. The specific formula is as follows:

[0039] In the formula, the desired heading angle is expressed as: , This represents the expected heading angle after smoothing. This represents the derivative of the expected heading angle after smoothing. Represents the velocity factor. Represents the filter factor. The fhan function represents the fastest control synthesis function, which has the function of quickly eliminating high-frequency oscillations. The fhan function is defined as follows:

[0040] 4. The extended state observer is used to observe the system's state variables and unknown disturbances other than wind. Inputs include: actual heading angle value. Expected control quantity Wind disturbance compensation The output includes: the estimated value of the actual heading angle, the estimated value of the derivative of the actual heading angle, and the estimated values ​​of other unknown disturbances besides wind. The specific formulas are as follows:

[0041] In the formula, The observation error represents the actual heading angle and the estimated actual heading angle; This represents the estimated actual heading angle; The estimated value representing the differential of the actual heading angle; This indicates the estimated value of other unknown disturbances besides wind. , , Represents the gain of the extended state observer; This represents a continuous power function with a linear segment near the origin; , represent The degree of nonlinearity of the function, ; represent Width of the linear interval of the function The function is represented as:

[0042] 5. Nonlinear state error feedback unit: This unit uses state errors to perform nonlinear combinations to obtain the desired control quantity. The inputs to the nonlinear state error feedback unit include: the deviation between the smoothed desired heading angle and the predicted actual heading angle; and the deviation between the predicted derivative of the smoothed desired heading angle and the predicted derivative of the actual heading angle. The output is the desired control quantity. The formula is as follows:

[0043] In the formula, Represents the proportionality coefficient. Represents the differential coefficient; , and This represents an adjustable parameter; To smooth out the deviation between the expected heading angle and the actual heading angle estimate; This is to smooth out the deviation between the estimated value of the derivative of the expected heading angle and the estimated value of the derivative of the actual heading angle after smoothing.

[0044] 6. The inputs to the disturbance compensation unit include: the desired control quantity. Wind disturbance compensation And other unknown disturbance estimates The output control parameters for the unmanned surface vessel are calculated using the following formula:

[0045] In the formula, This represents the equivalent input gain.

[0046] Example like Figure 2 As shown, to verify the performance of the proposed adaptive wind-feedback unmanned surface vessel (USV) heading control system based on active disturbance rejection, it was applied to USV path tracking, and hardware-in-the-loop simulation was performed. The simulation platform includes a remote control platform computer and the USV control system, with MQTT transparent communication between the remote control platform and the USV control system via a 4G network.

[0047] The unmanned surface vessel (USV) control system includes: a main control box, an ultra-wideband positioning device, an ultrasonic anemometer, and a propulsion system. The ultrasonic anemometer is responsible for collecting relative wind speed data. and relative wind direction The ultra-wideband positioning device is responsible for collecting the location information of the unmanned vessel. The collected information is transmitted to the main control box via the CAN bus. The main control box and the 4G-DTU communicate via serial port. The main control box displays the collected relative wind speed... relative wind direction Location information The data is transmitted to the MQTT server via the 4G network through the 4G-DTU. The remote control platform receives the data from the server and inputs it into the heading control system for real-time simulation to output control torque. The server sends control torque to the unmanned vessel, which then controls its course by changing the rudder angle of the propulsion device.

[0048] Figure 3 This is a comparison chart of the heading angle tracking performance of the unmanned surface vessel. 1 represents the heading angle tracking effect of the heading control system of this method. 2 represents the heading angle tracking performance of the active disturbance rejection heading control system without adaptive wind feedforward. It can be seen that the heading control system of this unmanned vessel has high robustness in complex wind fields, and its wind disturbance resistance is better than that of the active disturbance rejection heading control system without adaptive wind feedforward. It also has a smaller heading angle tracking error and can track the desired heading angle more accurately to complete the heading control task.

[0049] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A self-adaptive wind feedforward unmanned ship course control system based on active disturbance rejection, characterized in that, The utility model relates to an unmanned ship control system, including: An ultrasonic anemometer, a wind feedforward module, a gradient descent module and a disturbance-observer-based control module; The disturbance-observer-based control module includes a tracking differentiator, an extended state observer, a nonlinear state error feedback unit and a disturbance compensation unit; The ultrasonic anemometer collects relative wind speed and relative wind direction; The gradient descent module outputs compensation gain based on the difference between actual heading angle and desired heading angle; The wind feedforward module receives relative wind speed, relative wind direction and compensation gain and outputs wind disturbance compensation; The extended state observer outputs actual heading angle estimate, actual heading angle differential estimate and other unknown disturbance estimate; The nonlinear state error feedback unit uses state error for nonlinear combination to obtain desired control amount; The disturbance compensation unit determines unmanned ship control amount based on desired control amount, wind disturbance compensation and other unknown disturbance estimate.

2. The system of claim 1, wherein, The gradient descent module outputs compensation gain based on the difference between actual heading angle and desired heading angle, including: wherein represents a learning; represents a time step; represents a gain at a current time, represents a gain at a previous time; , is a difference between an actual course angle and a desired course angle, is an actual course angle, is a desired course angle; is a relative wind speed, is a relative wind direction; represents an air density; represents a waterplane area; represents a ship length.

3. The system of claim 2, wherein, The wind feedforward module receives relative wind speed, relative wind direction and compensation gain and outputs wind disturbance compensation, including: wherein, is the relative wind speed, is the relative wind direction; represents the air density; represents the waterplane area above the waterline; represents the length of the ship; is the compensation gain; is the wind disturbance compensation.

4. The system of claim 1, wherein, The extended state observer outputs actual heading angle estimate, actual heading angle differential estimate and other unknown disturbance estimate, including: This represents a continuous power function with a linear segment near the origin; , represent The degree of nonlinearity of the function, ; represent Width of the linear interval of the function The function is represented as: wherein represents the observation error of the actual course angle and the actual course angle estimate; represents the actual course angle estimate; represents the actual course angle differential estimate; represents the other unknown disturbance estimate; , , represents the extended state observer gain.

5. The system of claim 1, wherein, The nonlinear state error feedback unit uses state error for nonlinear combination to obtain desired control amount, including: wherein represents a proportional coefficient, represents a differential coefficient; , and represents an adjustable parameter; is a deviation of the smoothed expected heading angle from the estimated value of the actual heading angle; is a deviation of the smoothed differential of the expected heading angle from the estimated value of the differential of the actual heading angle; represents an expected control amount.

6. The system of claim 1, wherein, The tracking differentiator calculates smoothed desired heading angle and smoothed desired heading angle differential, including: The fhan function, representing the fastest control synthesis function, is defined as follows: where the desired heading angle is represented as , represents the smoothed desired heading angle, represents the derivative of the smoothed desired heading angle; represents the speed factor, represents the filter factor.

7. The system of claim 1, wherein, The disturbance compensation unit determines unmanned ship control amount based on desired control amount, wind disturbance compensation and other unknown disturbance estimate, including: wherein represents an equivalent input gain; represents a control amount of the unmanned ship; represents a desired control amount; represents an other unknown disturbance estimate; is a wind disturbance compensation.

Images