A two-stage hybrid propulsion amphibious unmanned platform active disturbance rejection control method and system

Abstract

The present invention provides a two-stage hybrid propulsion amphibious unmanned platform self-disturbance rejection control method and system, belonging to the field of amphibious unmanned platforms. In order to solve the problem that the existing self-disturbance rejection control method is difficult to track the expected speed or expected depth and the data processing speed is slow when it is applied to the control of amphibious unmanned platforms. Compared with the traditional control method, the present method does not require high precision of the control object and has strong anti-interference ability. At the same time, it can adjust the control parameters of the self-disturbance rejection controller, change the nonlinear error feedback control to a linear form, and use a linear extended state observer to well estimate the external interference of the amphibious unmanned platform and the internal disturbance of the system to compensate for it at the control quantity, which reduces the difficulty of parameter setting, has stronger adaptability and robustness, and is easy to operate; the sliding window filtering algorithm is used to filter the sensor data, filter out the noise interference of the signal, make the data fluctuation smoother, and speed up the data processing.

Description

Technical Field

[0001] The present invention relates to the technical field of amphibious unmanned platforms, and in particular to an automatic anti-disturbance control method and system for an amphibious unmanned platform with two-stage hybrid propulsion. Background Art

[0002] With the development of the global economy and the need for national defense, the vigorous development of the marine economy, the exploitation of marine resources, and the strengthening of marine armaments have become a general trend. However, with the deepening of marine development and the increasing demand for armaments, there is a need for amphibious unmanned vehicles that can operate in inaccessible or hazardous environments, carry a variety of functional equipment, and be able to navigate underwater, crawl underwater, and move on land to accomplish tasks such as marine environmental observation, seafloor topography mapping, underwater rescue, unmanned island landing operations, and military reconnaissance. Currently, most amphibious unmanned vehicles in various countries are still in the development and testing stage and have not yet been tested in actual combat. Their flexibility, autonomy, and coordination capabilities do not yet meet the requirements of amphibious operations, leaving significant room for technological development. Amphibious vehicles based on biomimetic and paddlewheel systems suffer from poor obstacle clearance, prone to slippage, and poor scalability, environmental adaptability, and maneuverability.

[0003] A multi-functional unmanned transport platform should possess advantages such as scalability, versatility, good maneuverability, rapid response, and reasonable price. This invention, based on an unmanned underwater vehicle (UUV), is based on practicality and versatility. The goal is to design a multi-functional unmanned transport platform with a certain carrying capacity, data collection capabilities, the ability to conduct autonomous unmanned movement, and effective human-machine interaction. This allows operators to monitor the platform's status in real time and promptly adjust mission planning to improve mission success rates.

[0004] Active disturbance rejection control (ADRC) technology is an emerging control theory that aims to solve the problem of disturbance suppression in traditional control systems. ADRC technology was proposed by Han Jingqing of the Chinese Academy of Sciences, and has since gradually developed and been widely used. ADRC technology has strong anti-interference capabilities and robustness, and can effectively cope with various disturbances inside and outside the system. It is one of the current research hotspots in the field of control. Traditional control systems often find it difficult to effectively cope with various disturbances when faced with complex systems, especially disturbances in nonlinear and time-varying systems. Traditional linear control methods such as PID controllers often require precise mathematical models and parameter adjustments. In actual engineering, it is often difficult to obtain an accurate mathematical model of the system. Therefore, it is easily affected by external disturbances, resulting in a decrease in system performance. In order to overcome this problem, active disturbance rejection control technology came into being.

[0005] The core concept of Active Disturbance Rejection Control (ADRC) is to estimate disturbances in the system by introducing an observer, treating them as part of the system, and incorporating compensation mechanisms into the controller, enabling the system to actively suppress disturbances. Compared to traditional control methods, ADRC does not require an accurate system model; it only requires measuring the system's inputs and outputs to achieve stable control, demonstrating its strong adaptability and robustness.

[0006] During the development of ADRC technology, researchers have continuously refined and optimized its theoretical framework and algorithms, proposing a series of improved methods and technologies. For example, the model-based ADRC method has conducted in-depth research on system modeling and parameter adjustment, improving the stability and performance of the system; the adaptive identification-based ADRC method (ADRC with Adaptive Identification) uses adaptive identification technology to update the system model in real time, improving the system's adaptability to parameter changes; the multi-model-based ADRC method (MM-ADRC) has achieved significant results in multi-model switching control and is suitable for the control of complex multi-modal systems. However, when existing ADRC methods are applied to the control of multi-habitat unmanned platforms, they have difficulty tracking the desired speed or depth and have slow data processing speeds. Summary of the Invention

[0007] The technical problems to be solved by the present invention are:

[0008] In order to solve the problems that the existing active disturbance rejection control method is difficult to track the expected speed or expected depth and the data processing speed is slow when applied to the control of multi-habitat unmanned platforms.

[0009] The present invention is to solve the above technical problems using the following technical solutions:

[0010] The present invention provides an active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion, comprising the following steps:

[0011] S100, establish a dynamic model of the multi-functional unmanned platform in water,

[0012] The hydraulic power and the control force of the actuator are expressed as follows:

[0013]

[0014] Among them, M A is the additional mass matrix, C A (V) is the hydrodynamic Coriolis matrix, D(V) is the hydrodynamic damping matrix, g(Θ) is the gravity buoyancy matrix, and τ is the control force matrix of the actuator;

[0015] in,

[0016]

[0017]

[0018]

[0019] The hydrodynamic damping is expressed as follows:

[0020] D(V)V=D l V+uD u V+D nl (V)V

[0021] Among them, D l is the linear damping matrix, D u is the damping matrix that increases with increasing longitudinal velocity, D nl (V) is the nonlinear damping matrix;

[0022] in,

[0023]

[0024]

[0025]

[0026] Assume that the mass of the amphibious unmanned platform immersed in water is W = mg, where g is the acceleration due to gravity; the buoyancy is B = ρg▽, where ρ is the water density and ▽ is the volume of the amphibious unmanned platform immersed in water; assume that the gravity and buoyancy are equal, that is, W = B, and the center of buoyancy of the amphibious unmanned platform is r B =[x B ,y B ,z B ] T , and y B =0, then we have the following vectors:

[0027]

[0028] The amphibious unmanned platform is an underactuated amphibious unmanned platform, and the longitudinal thrust, bow moment and pitch moment are expressed in vector form as follows:

[0029] τ=[τ u 000τ q τ r ] T

[0030] Among them, τ u is the longitudinal thrust, τ q is the pitch moment, τ r is the bow turning moment;

[0031] By simplifying the six-degree-of-freedom mathematical model, the horizontal plane mathematical model of the amphibious unmanned platform can be obtained, which can be expressed in the form of a differential equation system as follows:

[0032]

[0033] Where A = -d 22 v+(d 26 -uc 26 )r,B=(d 62 -uc 62 )vd 66 r+τ r ,and,

[0034]

[0035] S200. Establish an active disturbance rejection controller for a multi-habitat unmanned platform. The active disturbance rejection controller includes a tracking differentiator, a linear extended state observer, and a linear error feedback module. The tracking differentiator is used to arrange the transition process to smooth the sudden changes in the input signal. The extended state observer tracks the influence of the unknown parts of the model and the external unknown disturbances by establishing an extended state quantity. The linear error feedback module provides a control quantity to compensate for the disturbance, so that the multi-habitat unmanned platform adjusts its heading in real time, and outputs the current actual heading and re-inputs it into the linear extended state observer to continue the cyclic control.

[0036] Furthermore, in step S200, it includes:

[0037] S210, the tracking differentiator receives the set desired heading angle information ψ d , calculating an approximate value x1 of the desired heading angle and a differential signal x2 of the approximate heading angle according to the kinematic model of step S100;

[0038] S220, the linear extended state observer receives the actual heading angle information ψ and the input control variable u of the multi-habitable unmanned platform, calculates the estimated values ​​of the actual heading of the multi-habitable unmanned platform and its first-order differential, which are expressed as z1 and z2 respectively, and simultaneously calculates the total disturbance z3 suffered by the multi-habitable unmanned platform;

[0039] S230, calculating the deviation e1 between x1 obtained by the tracking differentiator and the observed variable z1 and the deviation e2 between the corresponding differential signals x2 and z2, and obtaining e1 = x1 - z1, e2 = x2 - z2;

[0040] S240, the linear error feedback module obtains a control component according to the error e1 and the error e2, and performs disturbance compensation on the control component using the estimated value z3 of the unknown disturbance to obtain the control variable u of the multi-habitat unmanned platform;

[0041] S250: After receiving the control variable u, the multi-habitat unmanned platform adjusts its heading in real time, and outputs the current actual heading to the linear expansion state observer to continue the cyclic control.

[0042] Furthermore, in step S210, a sliding window filtering algorithm is used to filter the data sampled by the sensors in the navigation and positioning system, including:

[0043] Set a sampling window of length M and a filtering window of length N. After the sensor starts sampling, if the data, i.e., the number of sample points, is less than M, the data in the sampling window is accumulated and the average value is calculated as the filtered result; if the number of data sample points is greater than M, the sampling data before time M is discarded, and the data in the sampling window is bubble sorted, and the maximum and minimum values ​​are removed. The remaining M-2 data form the filtering window, satisfying M-2=N. The data in the filtering window is accumulated and the average value is calculated to obtain the filtered result.

[0044] Furthermore, when the amphibious unmanned platform is performing depth-fixed navigation, if the difference between the sum of the actual depth and the actual height minus the expected depth is less than 0.5 meters, the amphibious unmanned carrier platform will change from depth-fixed navigation to 0.5-meter height-fixed navigation.

[0045] Furthermore, the amphibious unmanned platform can achieve bow control of the amphibious unmanned platform by controlling the vertical rudder during high-speed navigation, including:

[0046] In rudder mode, incremental PID control is used. The error between the actual heading angle and the expected value is used as the input of the PID controller, and the output is the corresponding rudder angle deflection. The expression is as follows:

[0047] Δθ(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0048] Among them, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, e(k-2) is the heading angle error at the moment before that, and K p is the proportional gain term, K i is the integral gain term, K d is the differential gain term.

[0049] Furthermore, the in-situ steering of the multi-habitat unmanned platform is achieved through the main propulsion differential, using incremental PID control. The error between the actual heading angle and the expected value is used as the input of the PID controller, and the output is the resultant force, which is expressed as follows:

[0050] ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0051] Wherein, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, and e(k-2) is the heading angle error at the previous moment; ΔF=|F1|+|F2|, where F1 is the thrust allocated to the left main thruster (501), F2 is the thrust allocated to the right main thruster (502), and F1+F2=0.

[0052] Furthermore, the speed control of the multi-habitat unmanned platform includes speed maintenance and maneuvering speed change. Incremental PID control is used. The error between the actual speed and the expected value is used as the input of the PID controller, and the output is the resultant force. Its expression is as follows:

[0053] ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0054] Wherein, e(k) is the speed error at the current moment, e(k-1) is the speed error at the previous moment, and e(k-2) is the speed error at the previous moment; ΔF=|F1|+|F2|, F1 is the thrust allocated to the left main thruster (501), and F2 is the thrust allocated to the right main thruster (502).

[0055] Furthermore, in the horizontal rudder mode, the depth control of the multi-roofed unmanned platform is achieved through the vertical thrusters, horizontal rudders and main thrusters, using an incremental PID control algorithm, with the depth error as the input of the PID controller and the horizontal rudder angle as the output;

[0056] Combined with the line of sight navigation method, the 10m in front of the multi-habitat unmanned platform's own X direction position is used as the desired X direction. d Direction position, the desired depth position is Z d , we get the displacement error equation:

[0057] x e =xx d ,z e =zz d

[0058] The desired pitch angle is:

[0059] θ d =tan -1((zz d ) / (xx d )).

[0060] The present invention provides an automatic disturbance rejection control system for an amphibious unmanned platform with two-stage hybrid propulsion. The system has a program module corresponding to the above steps, and executes the steps in the automatic disturbance rejection control method of the two-stage hybrid propulsion amphibious unmanned platform when running.

[0061] The present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program is configured to implement the steps of a two-stage hybrid propulsion amphibious unmanned platform self-disturbance rejection control method when called by a processor.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] The present invention discloses an automatic disturbance rejection control method and system for an amphibious multi-purpose unmanned platform with two-stage hybrid propulsion. Compared with traditional control methods, the method has low precision requirements for the controlled object and strong anti-interference ability. At the same time, the method can adjust the control parameters of the automatic disturbance rejection controller, and change the nonlinear error feedback control to a linear form. The linear extended state observer is used to well estimate the external interference of the amphibious multi-purpose unmanned platform and the internal disturbance of the system, and compensate for it at the control quantity, which reduces the difficulty of parameter setting, enhances adaptability and robustness, and is easy to operate. The sliding window filtering algorithm is used to filter the sensor data, filter out the noise interference of the signal, make the data fluctuation smoother, and speed up the data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a front view of the multi-habitat unmanned platform in an embodiment of the present invention;

[0065] Figure 2 1. A top view of a multi-habitat unmanned platform according to an embodiment of the present invention;

[0066] Figure 3 Schematic diagram of the structure of the stern direction of the multi-habitat unmanned platform in an embodiment of the present invention;

[0067] Figure 4 Schematic diagram of the active disturbance rejection control system structure of a multi-habitat unmanned platform in an embodiment of the present invention;

[0068] Figure 5 This is a parameter setting diagram for an amphibious unmanned platform according to an embodiment of the present invention;

[0069] Figure 6 Schematic diagram of the main thrust differential left turn control thrust distribution in an embodiment of the present invention;

[0070] Figure 7 Schematic diagram of speed control thrust distribution of a multi-habitat unmanned platform in an embodiment of the present invention;

[0071] Figure 8 Schematic diagram of the depth control thrust distribution of the multi-habitat unmanned platform in an embodiment of the present invention.

[0072] Description of reference numerals:

[0073] 2. Track chassis; 3. Battery compartment; 4. Electronic compartment; 5. External frame; 6. Buoyancy block; 201. Driving wheel; 202. Load wheel; 203. Guide wheel; 204. Track; 501. Left main thruster; 502. Right main thruster; 503. Stern vertical thruster; 504. Stern horizontal thruster; 505. Bow vertical thruster; 506. Bow horizontal thruster; 508. Doppler sensor. DETAILED DESCRIPTION

[0074] In the description of the present invention, it should be noted that the terminology in each embodiment, such as "up", "down", "front", "back", "left", "right", etc., which indicate directions, are only for simplifying the description of the positional relationship based on the drawings in the specification, and do not mean that the referred elements and devices must be operated in accordance with the specific directions and defined operations and methods and structures in the specification. Such directional nouns do not constitute a limitation to the present invention.

[0075] In the description of the present invention, it should be noted that the terms "first," "second," and "third" mentioned in the embodiments of the present invention are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly specifying the number of the technical features indicated. Therefore, a feature specified as "first," "second," or "third" may explicitly or implicitly include one or more of such features.

[0076] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0077] Specific implementation plan 1: Combined Figures 1 to 4 As shown, the present invention provides an active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion, comprising the following steps:

[0078] S100, establish a dynamic model of an amphibious unmanned platform in water, including:

[0079] The hydraulic power and the control force of the actuator are expressed as follows:

[0080]

[0081] Among them, M A is the additional mass matrix, C A(V) is the hydrodynamic Coriolis matrix, D(V) is the hydrodynamic damping matrix, g(Θ) is the gravity buoyancy matrix, and τ is the control force matrix of the actuator;

[0082] in,

[0083]

[0084]

[0085]

[0086] Establishing an accurate six-degree-of-freedom hydrodynamic drag and torque matrix for an amphibious unmanned platform is a very challenging task. Finding hydrodynamic coefficients that meet all speed requirements is nearly impossible, especially for amphibious unmanned platforms operating within a wide speed range. Furthermore, when the longitudinal speed of the amphibious unmanned platform is relatively high, the nonlinearity of the hydrodynamics becomes very pronounced. Simultaneously, the coupling between the various degrees of freedom of the amphibious unmanned platform also increases. The hydrodynamic coefficients used in the present invention are derived from empirical formulas and tank experiments, and their accuracy is guaranteed when the amphibious unmanned platform operates at a speed of 2-4 m / s.

[0087] Combined with the above discussion, for amphibious unmanned platforms, the hydrodynamic damping can be expressed as follows:

[0088] D(V)V=D l V+uD u V+D nl (V)V

[0089] Among them, D l is the linear damping matrix, D u is the damping matrix that increases with increasing longitudinal velocity, D nl (V) is the nonlinear damping matrix;

[0090] in,

[0091]

[0092]

[0093]

[0094] Assume that the mass of the amphibious unmanned platform immersed in water is W = mg, where g is the acceleration due to gravity; the buoyancy is B = ρg▽, where ρ is the water density and ▽ is the volume of the amphibious unmanned platform immersed in water; assume that the gravity and buoyancy are equal, that is, W = B, and the center of buoyancy of the amphibious unmanned platform is r B =[x B ,y B ,z B ]T , and y B =0, then we have the following vectors:

[0095]

[0096] The amphibious unmanned platform of the present invention is an underactuated amphibious unmanned platform, and its actuator can only provide longitudinal thrust, bow moment and pitch moment, which can be expressed in vector form as follows:

[0097] τ=[τ u 000τ q τ r ] T

[0098] Among them, τ u is the longitudinal thrust, τ q is the pitch moment, τ r is the bow turning moment;

[0099] By simplifying the six-degree-of-freedom mathematical model, the horizontal plane mathematical model of the amphibious unmanned platform can be obtained, which can be expressed in the form of a differential equation system as follows:

[0100]

[0101] Where A = -d 22 v+(d 26 -uc 26 )r,B=(d 62 -uc 62 )vd 66 r+τ r ,and,

[0102]

[0103] The parameters of the amphibious unmanned platform involved in the present invention are as follows: Figure 5 As shown;

[0104] S200. Establish an active disturbance rejection controller for the amphibious unmanned platform. The active disturbance rejection controller includes a tracking differentiator, a linear extended state observer, and a linear error feedback module. The tracking differentiator is used to arrange the transition process, which can smooth the sudden change of the input signal, alleviating the contradiction between rapidity and overshoot in PID control technology. It can also extract the differential signal of the input signal to avoid noise amplification. The extended state observer is used to resolve the combined impact of the unknown part of the model and the external unknown disturbance on the control object. The impact of the unknown part of the model and the external unknown disturbance is tracked by establishing an extended state variable. Then, a control variable is given to compensate for these disturbances, and the control object is converted into a common integral series type control object. The purpose of establishing the extended state observer is to observe the extended state variable to estimate the unknown disturbance and the unmodeled part of the control object, realize feedback linearization of the dynamic system, and convert the control object into an integral series type.

[0105] Specifically include:

[0106] S210, the tracking differentiator receives the set desired heading angle information ψ d , calculate the approximate value x1 of the desired heading angle and the differential signal x2 of the approximate heading angle;

[0107] S220. The linear extended state observer receives the actual heading angle information ψ and the input control variable u of the amphibious unmanned platform, calculates the estimated values ​​of the actual heading of the amphibious unmanned platform and its first-order differential, which are expressed as z1 and z2 respectively, and simultaneously calculates the total disturbance z3 to the amphibious unmanned platform. The total disturbance z3 includes the external environmental force called external disturbance and the system internal disturbance caused by model inaccuracy or model parameter change.

[0108] S230, calculating the deviation e1 between x1 obtained by the tracking differentiator and the observed variable z1 and the deviation e2 between the corresponding differential signals x2 and z2, and obtaining e1 = x1 - z1, e2 = x2 - z2;

[0109] S240, the linear error feedback module obtains a control component according to the error e1 and the error e2, and performs disturbance compensation on the control component using the estimated value z3 of the unknown disturbance, thereby obtaining a control variable u of the amphibious unmanned platform;

[0110] S250: After receiving the control variable u, the multi-habitat unmanned platform adjusts its heading in real time, and outputs the current actual heading to the linear expansion state observer to continue the cyclic control.

[0111] Preferably, in step S210, a sliding window filtering algorithm is used to filter the data sampled by the sensors in the navigation and positioning system. A sampling window of length M and a filtering window of length N are set. After the sensor starts sampling, if the number of data sample points does not fill the sampling window, that is, the number of sample points is less than M, the data in the sampling window are accumulated and averaged as the filtered result. If the data sample points have filled the sampling window, that is, the number of sample points is greater than M, the sampling data before time M are discarded, and only the M sampling data obtained in the most recent M time moments are considered to form the sampling window. Then, after bubble sorting the data in the sampling window, the maximum and minimum values ​​therein are removed, and the remaining M-2 data form the filtering window, that is, satisfying M-2=N. The data in the filtering window are accumulated and averaged to obtain the filtered result. The sensor data processed by the sliding window filtering algorithm filters out noise interference, and the data fluctuation is more stable and smooth. Since the data before time M is discarded each time a new data is sampled, a new average value can be calculated for each sampling, which speeds up data processing.

[0112] Specific implementation plan 2: When the multi-purpose unmanned vehicle is performing depth-fixing navigation, it can adaptively switch between depth-fixing and altitude-fixing modes to ensure the safety of the multi-purpose unmanned vehicle; specifically, it includes:

[0113] During depth-fixed navigation, when the actual depth of the amphibious unmanned platform is greater than the expected depth, in order to avoid collision between the amphibious unmanned carrier platform and the seabed due to sudden changes in the seabed topography, if the difference between the sum of the actual depth and the actual height minus the expected depth is less than 0.5 meters, the amphibious unmanned carrier platform will change from depth-fixed navigation to 0.5-meter height-fixed navigation to ensure its safety during operation and avoid collision accidents caused by sudden changes in the seabed topography.

[0114] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0115] Specific implementation plan three: The amphibious unmanned platform controls the heading of the amphibious unmanned platform by controlling the vertical rudder during high-speed navigation. If the vertical rudder is used for steering, the amphibious unmanned platform needs to reach a certain speed to generate a suitable rudder effect. The heading control method of the amphibious unmanned platform at high speed is as follows:

[0116] In rudder mode, the amphibious unmanned platform lacks horizontal thrust and is an underactuated system. To control the heading of the vehicle, incremental PID control is used. The error between the actual heading angle and the desired value is used as the input of the PID controller, and the output is the corresponding rudder angle deflection. The expression is as follows:

[0117] Δθ(k)=Kp (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0118] Among them, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, e(k-2) is the heading angle error at the moment before that, and K p is the proportional gain term, K i is the integral gain term, K d is the differential gain term.

[0119] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0120] Specific implementation plan four: combined Figure 6 As shown in the figure, the in-situ steering of the multi-purpose unmanned platform is achieved through the main thrust differential. Incremental PID control is also used. The error between the actual heading angle and the expected value is used as input. The output is the resultant force that needs to be distributed between the two main thrust differentials. Then, reasonable torque is generated through thrust distribution:

[0121] In the main propulsion differential mode, the multi-route unmanned platform is also an underactuated system. To control the heading of the vehicle, incremental PID control is used. The error between the actual heading angle and the desired value is used as the input of the PID controller, and the output is the resultant force. Its expression is as follows:

[0122] ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0123] Among them, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, and e(k-2) is the heading angle error at the moment before that;

[0124] ΔF=|F1|+|F2|, where F1 is the thrust allocated to the left main thruster 501, F2 is the thrust allocated to the right main thruster 502, and F1+F2=0.

[0125] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0126] Specific implementation plan five: combined Figure 7As shown in the figure, the speed control of the multi-habitat unmanned platform mainly tests the speed control capability of the multi-habitat unmanned platform, including speed maintenance and maneuvering speed change capability. The actual speed and expected speed error of the multi-habitat unmanned platform are used as input, and the output is the resultant force that needs to be distributed between the two main thrusters.

[0127] In the multi-habitat unmanned platform speed control mode, the multi-habitat unmanned platform is also an underactuated system. In order to control the speed of the vehicle, incremental PID control is used. The error between the actual speed and the expected value is used as the input of the PID controller, and the output is the resultant force. Its expression is as follows:

[0128] ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2))

[0129] Wherein, e(k) is the speed error at the current moment, e(k-1) is the speed error at the previous moment, and e(k-2) is the speed error at the previous moment; ΔF=|F1|+|F2|, where F1 is the thrust allocated to the left main thruster 501, and F2 is the thrust allocated to the right main thruster 502.

[0130] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0131] Specific implementation plan six: combined Figure 8 As shown in the figure, the depth control of the amphibious unmanned platform is completed by the vertical thrusters, horizontal rudders and main thrusters. When the amphibious unmanned platform is sailing at a high speed, the horizontal rudders produce a steering effect, thereby controlling the amphibious unmanned platform to dive. When the amphibious unmanned platform needs to dive to a fixed depth, the vertical thrusters are required to complete the depth control.

[0132] In the horizontal rudder mode, if the amphibious unmanned platform wants to float or dive, the horizontal rudder is used to control the pitch angle of the amphibious unmanned platform, and the propulsion of the main thrust is coordinated to achieve depth control. The incremental PID control algorithm is used, and the depth error is used as the input of the PID controller to output the horizontal rudder angle.

[0133] Combined with the line of sight navigation method, the 10m in front of the multi-habitat unmanned platform's own X direction position is used as the desired X direction. d Direction position, the desired depth position is Z d , from which the displacement error equation can be obtained:

[0134] x e =xx d ,z e =zz d

[0135] According to the line of sight method, the desired pitch angle can be obtained, and its expression is as follows:

[0136] θ d =tan -1 ((zz d ) / (xx d )).

[0137] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0138] Specific implementation plan seven: The crawling of the amphibious unmanned platform to the target needs to be completed by the tracks. During the underwater crawling process, the steering of the amphibious unmanned platform is achieved by controlling the differential speed of the tracks through PID, and the acceleration, deceleration, forward and backward maneuvers are achieved by controlling the duty cycle of the track motor; in the underwater environment, the thrusters and tracks are combined to control the amphibious unmanned platform to crawl and maneuver close to the bottom of the water, and the pressure of the amphibious unmanned platform and the bottom of the water is increased by controlling the vertical thrusters to apply downward thrust, and the main thrust is controlled to assist the tracks in crawling in the water.

[0139] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0140] The upper platform of the multi-functional unmanned platform includes a closed electronic cabin 4 and an external frame 5 outside the electronic cabin 4. The electronic cabin 4 is arranged at the center of the external frame 5. The electronic cabin 4 is equipped with a main control system, a navigation and positioning system, a communication system, and a power control system. The external frame 5 is equipped with a water propulsion system. A land power unit is provided below the upper platform.

[0141] The underwater propulsion system includes a left main propeller 501 and a right main propeller 502 symmetrically arranged on both sides of the central axis of the stern of the amphibious unmanned platform, and also includes a stern vertical propeller 503, a stern horizontal propeller 504, a bow vertical propeller 505 and a bow horizontal propeller 506. The propellers of the left main propeller 501 and the right main propeller 502 are horizontally facing the rear of the amphibious unmanned platform; the propeller of the stern vertical propeller 503 is arranged at the bottom of the external frame 5 and the propeller of the stern vertical propeller 503 is facing the tracked chassis 2 of the land power unit; the stern horizontal propeller 504 and The bow horizontal thrusters 506 are parallel to each other and are located on both sides of the bow and stern of the amphibious unmanned platform; the propeller of the bow vertical thruster 505 is oriented perpendicular to the port side; the propeller of the bow horizontal thruster 506 is oriented perpendicular to the starboard side. The coordinated use of the six thrusters can realize the amphibious unmanned platform's maneuvers of surfacing, diving, and turning. The thrusters are supplied with power and control signal power by the electronic compartment 4. The thrusters are 48V-powered watertight brushless DC motors that can switch forward and reverse at any time, adopt closed-loop control, provide real-time feedback of thrust data, and have an overcurrent protection mechanism.

[0142] The land power system includes a crawler chassis 2 and two crawler assemblies arranged below the crawler chassis 2, each crawler assembly includes a driving wheel 201, a guide wheel 203, a load wheel 202, a crawler 204 and a crawler motor, and several load wheels 202 are evenly arranged between the driving wheel 201 and the guide wheel 203, and the driving wheel 201 and the guide wheel 203 are both located obliquely above the load wheels 202 at both ends, and the crawler 204 is wrapped around the outside of the driving wheel 201, the guide wheel 203 and several load wheels 202, the output end of the crawler motor is connected to the driving wheel 201, and the driving wheel 201, the guide wheel 203 and the load wheel 202 are all rotatably connected to the crawler chassis 2; the amphibious unmanned platform realizes forward and backward movement and left and right turning on land through the forward and backward rotation and differential speed of the two crawler assemblies;

[0143] The main control system includes a motion computer, a mission computer, and a guidance computer. The motion computer is used to sample and process sensor data and feed the sensor data back to the host computer; the mission computer is used to plan tasks, execute motion algorithms, record information about the multi-terrestrial unmanned platform, and perform network communications and fault monitoring; the guidance computer is used to give guidance instructions, including desired heading, desired speed, and desired depth; the motion control algorithm of the multi-terrestrial unmanned platform is run by the main control system, and the instructions generated after calculation can be transmitted to the underwater propulsion system or the land power system, thereby controlling the movement of the multi-terrestrial unmanned platform in water, on land, or underwater;

[0144] The navigation and positioning system includes a GPS module in the electronic cabin 4, an inertial combined sensor module (including an inertial navigation sensor and a Doppler sensor 508 (DVL) located outside the cabin) and a depth gauge; it is used to locate and calculate the attitude of the multi-habitat unmanned platform, and realize integrated navigation through the fusion of multiple sensors;

[0145] The communication system includes a WiFi communication module and a wireless communication module, which are used to connect to the host computer through the communication system. The communication mode can be adjusted according to the environment. The hybrid communication mode can ensure that the multi-habitat unmanned carrier platform can communicate reliably over a large range.

[0146] The power control system includes a main single-chip microcomputer control board and a relay matrix board. Each pin of the main single-chip microcomputer control board controls a relay switch, thereby controlling the power supply of each module. The relay board is equipped with a voltage divider module, which divides the 48V battery voltage into three circuits: 48V main power, 24V auxiliary power, and 5-12V control power, and controls them in stages through relays. The rated capacity of the battery is 300Wh, and the total battery capacity is 600Wh. The 48V main power is responsible for powering the six thrusters, the 24V auxiliary power is responsible for powering the high-power sensors and computer, and the 5-12V control is responsible for powering the low-power sensors and single-chip microcomputer. The relay and current are sampled and processed by the single-chip microcomputer board and then sent to the mission computer. The electronic compartment 4 uses dual power supply to avoid the current reduction caused by the operation of multiple motors, which may affect the overall performance of the multi-functional unmanned platform. The power supply is a rechargeable high-energy battery, which is installed in the battery compartment 3 provided at the stern of the external frame 5. The power control system is responsible for power supply management of all power devices and monitoring the battery voltage and capacity.

[0147] Preferably, a plurality of buoyancy blocks 6 are provided above the crawler chassis 2. The buoyancy blocks 6 are foam-type buoyancy materials. The amphibious unmanned platform can be balanced by the buoyancy blocks 6. After balancing, the center of gravity and the center of buoyancy of the amphibious unmanned platform are located at the geometric center position, and the overall balancing is slightly negative buoyancy, which is conducive to the amphibious unmanned platform to travel on the bottom.

[0148] Preferably, the track motor can be powered by a battery on the track chassis 2. The track chassis 2 is provided with a watertight control cabin. The control cabin contains a track microcontroller that directly controls the track motor and communicates with the main control system through a wired network connection. The chassis control cabin is provided with a water cooling circulation system to cool the control cabin. The track motor is also equipped with an encoder to measure the movement speed of the track 204 by measuring the motor speed within a certain period of time.

[0149] Preferably, the electronic cabin 4 is a cylindrical stainless steel shell structure, which can effectively conduct heat in an underwater environment, and a heat dissipation module is provided in the electronic cabin 4; end covers with watertight piercing parts are provided at both ends of the shell of the electronic cabin 4; there are rubber watertight interfaces on the end covers, which can establish electrical and signal connections with external thrusters, sensors, and chassis components through watertight cables; the electronic cabin 4 uses sealing rings and vacuum treatment to ensure internal sealing.

[0150] Preferably, for the controlled system, a PID controller based on radial basis function network tuning is set. The radial basis function network is synonymous with RBF radial basis function network, RBF network and RBF neural network. The PID adaptive control based on the RBF radial basis function network is an effective control method. It combines the simplicity of the PID controller and the nonlinear approximation ability of the RBF network, and can better adapt to the control requirements of various complex systems.

[0151] Since traditional PID control requires constant parameter adjustment, it is difficult to achieve optimal results if the parameters are unreasonable. It takes a long time to achieve the optimal effect. Therefore, on the basis of single neuron adaptive PID control, a PID controller based on RBF neural network tuning is used; compared with the single neuron adaptive PID controller, its advantage is that it is nonlinear; the control effect of the PID controller depends on the three control functions of proportion, integration and differentiation, forming a relationship of mutual cooperation and mutual restraint in the control quantity, but the effect is not necessarily linear, so the use of nonlinear RBF can achieve better results; the arbitrary nonlinear expression ability of RBF neural network can realize PID control with the best combination by learning the system performance;

[0152] The radial basis function neural network is a three-layer feedforward network with a single hidden layer. The input to output is nonlinear, while the mapping from the hidden layer space to the output is linear, which speeds up the learning speed and avoids the local minimum problem. Its structure is as follows: Figure 4 As shown, in the RBF network structure, X=[x1,x2,...,x n ] T Assume that the network input vector is the radial basis vector of the hidden layer node of the RBF neural network H=[h1,h2...,h m ] T , where h j is the Gaussian basis function:

[0153]

[0154] The center vector of the jth node in the hidden layer of the RBF neural network is C j =[c j1 ,c j2 ,...,c jm ] T , let the base width vector of the hidden layer node of the RBF neural network be B=[b1,b2,....,b m ] T , where b j is the basis width parameter of hidden layer node j, and is a number greater than 0; the weight vector of the network is w=[w1,w2,...,w n ] T , the identified network output is:

[0155] y mout =w1h1+w2h2+...+w m h m

[0156] Combine Figure 5The block diagram of the PID controller based on RBF network tuning is shown in the figure. The parameter adjustment algorithm in the network is NNI parameter adjustment. Assume that the theoretical output of the identified system at time k is y out (k), the output of network identification is y mout (k), then the identification system performance index function is:

[0157]

[0158] The iterative algorithm for obtaining the output weight, node center and node base width parameters by the gradient descent method is as follows:

[0159] The output weight expression is as follows:

[0160] w j (k) = w j (k-1)+η(y out (k)-y mout (k))h j +α(w j (k-1)-w j (k-2))+β(w j (k-2)-w j (k-3))

[0161] Where η is the learning rate; α and β are momentum factors, both ranging from (0, 1);

[0162] Node base width change Δb j for:

[0163]

[0164] Node base width parameter b j for:

[0165] b j (k) = b j (k-1)+ηΔb j +α(b j (k-1)-b j (k-2))+β(b j (k-2)-b j (k-3))

[0166] Node center change Δc ji for:

[0167]

[0168] Node center c j for:

[0169] c ji =c ji(k-1)+ηΔc ji +α(c ji (k-1)-c ji (k-2))+β(c ji (k-2)-c ji (k-3))

[0170] Where cj = [c11, c12, c13…c1m] is the center vector value of the jth hidden layer neuron;

[0171] The PID parameter adjustment algorithm is as follows:

[0172] The control error is:

[0173] error(k)=u(k)-yout(k)

[0174] Among them, u(k) is the input of the controlled system, yout(k) is the output of the controlled system;

[0175] The 3 PID inputs are:

[0176] xc(1)=error(k)-error(k-1)

[0177] xc(2)=error(k)

[0178] xc(3)=error(k)-2error(k-1)+error(k-2)

[0179] The neural network tuning index is:

[0180]

[0181] By gradient descent method, we get three parameters K p ,K I ,K D The adjustment amount is:

[0182]

[0183]

[0184]

[0185] in, K is the Jacobian information of the controlled object, that is, the sensitivity of the object's output to the change of the control input, which can be identified by the RBF neural network; p is the proportional gain term, K i is the integral gain term, K d is the differential gain term;

[0186]

[0187] The expected depth and the actual depth of the multi-habitat unmanned platform are used as errors. The RBF network structure is 3-6-1. The three parameters of network identification are the control input increment, the depth at the current moment, and the depth information at the previous moment. The initial proportional, integral, and differential parameters are set to: 0.01, 0.02, and 0.03. When the simulation ends, they become 436.5727, 60.9628, and 1117. It can be seen that the three parameters will be continuously adjusted during the simulation process.

[0188] Specific implementation scheme eight: The present invention provides a two-stage hybrid propulsion amphibious unmanned platform self-disturbance rejection control system, which has a program module corresponding to the above steps and executes the steps in the above-mentioned two-stage hybrid propulsion amphibious unmanned platform self-disturbance rejection control method during operation.

[0189] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0190] Specific implementation scheme nine: The present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the self-disturbance rejection control method of an amphibious unmanned platform with two-stage hybrid propulsion when called by a processor.

[0191] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0192] Although the present invention is disclosed as above, the scope of protection disclosed by the present invention is not limited thereto. Those skilled in the art of the present invention may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A two-stage hybrid propulsion amphibious unmanned platform active disturbance rejection control method, characterized in that: The following steps are involved: S100, establish a dynamic model of the multi-functional unmanned platform in water, The hydraulic power and the control force of the actuator are expressed as follows: Among them, M A is the additional mass matrix, C A (V) is the hydrodynamic Coriolis matrix, D(V) is the hydrodynamic damping matrix, g(Θ) is the gravity buoyancy matrix, and τ is the control force matrix of the actuator; in, The hydrodynamic damping is expressed as follows: D(V)V=D l V+uD u V+D nl (V)V Among them, D l is the linear damping matrix, D u is the damping matrix that increases with increasing longitudinal velocity, D nl (V) is the nonlinear damping matrix; in, Assume that the mass of the amphibious unmanned platform immersed in water is W = mg, where g is the acceleration due to gravity; the buoyancy is ρ is the density of water, is the volume of the amphibious unmanned platform immersed in water; assuming that gravity and buoyancy are equal, that is, W = B, the buoyancy center of the amphibious unmanned platform is r B =[x B ,y B ,z B ] T , and y B =0, then we have the following vectors: The amphibious unmanned platform is an underactuated amphibious unmanned platform, and the longitudinal thrust, bow moment and pitch moment are expressed in vector form as follows: τ=[τ u 000t q t r ] T Among them, τ u is the longitudinal thrust, τ q is the pitch moment, τ r is the bow turning moment; By simplifying the six-degree-of-freedom mathematical model, the horizontal plane mathematical model of the amphibious unmanned platform can be obtained, which can be expressed in the form of a differential equation system as follows: Where A = -d 22 v+(d 26 -uc 26 )r,B=(d 62 -uc 62 )vd 66 r+τ r ,and, S200. Establish an active disturbance rejection controller for a multi-habitat unmanned platform. The active disturbance rejection controller includes a tracking differentiator, a linear extended state observer, and a linear error feedback module. The tracking differentiator is used to arrange the transition process to smooth the sudden changes in the input signal. The extended state observer tracks the influence of the unknown parts of the model and the external unknown disturbances by establishing an extended state quantity. The linear error feedback module provides a control quantity to compensate for the disturbance, so that the multi-habitat unmanned platform adjusts its heading in real time, and outputs the current actual heading and re-inputs it into the linear extended state observer to continue the cyclic control.

2. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 1 is characterized in that: In step S200, it includes: S210, the tracking differentiator receives the set desired heading angle information ψ d , calculating an approximate value x1 of the desired heading angle and a differential signal x2 of the approximate heading angle according to the kinematic model of step S100; S220, the linear extended state observer receives the actual heading angle information ψ and the input control variable u of the multi-habitable unmanned platform, calculates the estimated values ​​of the actual heading of the multi-habitable unmanned platform and its first-order differential, which are expressed as z1 and z2 respectively, and simultaneously calculates the total disturbance z3 suffered by the multi-habitable unmanned platform; S230, calculating the deviation e1 between x1 obtained by the tracking differentiator and the observed variable z1 and the deviation e2 between the corresponding differential signals x2 and z2, and obtaining e1 = x1 - z1, e2 = x2 - z2; S240, the linear error feedback module obtains a control component according to the error e1 and the error e2, and performs disturbance compensation on the control component using the estimated value z3 of the unknown disturbance to obtain the control variable u of the multi-habitat unmanned platform; S250: After receiving the control variable u, the multi-habitat unmanned platform adjusts its heading in real time, and outputs the current actual heading to the linear expansion state observer to continue the cyclic control.

3. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 2 is characterized by: In step S210, a sliding window filtering algorithm is used to filter the data sampled by the sensors in the navigation and positioning system, including: Set a sampling window of length M and a filtering window of length N. After the sensor starts sampling, if the data, i.e., the number of sample points, is less than M, the data in the sampling window is accumulated and the average value is calculated as the filtered result; if the number of data sample points is greater than M, the sampling data before time M is discarded, and the data in the sampling window is bubble sorted, and the maximum and minimum values ​​are removed. The remaining M-2 data form the filtering window, satisfying M-2=N. The data in the filtering window is accumulated and the average value is calculated to obtain the filtered result.

4. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 3 is characterized by: When the amphibious unmanned platform is performing depth-fixed navigation, if the difference between the sum of the actual depth and the actual altitude minus the expected depth is less than 0.5 meters, the amphibious unmanned platform will change from depth-fixed navigation to 0.5-meter altitude-fixed navigation.

5. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 4 is characterized by: The amphibious unmanned platform can achieve bow control of the amphibious unmanned platform by controlling the vertical rudder during high-speed navigation, including: In rudder mode, incremental PID control is used. The error between the actual heading angle and the expected value is used as the input of the PID controller, and the output is the corresponding rudder angle deflection. The expression is as follows: Δθ(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2)) Among them, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, e(k-2) is the heading angle error at the moment before that, and K p is the proportional gain term, K i is the integral gain term, K d is the differential gain term.

6. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 5 is characterized by: The in-situ steering of the multi-habitat unmanned platform is achieved through the main propulsion differential, using incremental PID control. The error between the actual heading angle and the expected value is used as the input of the PID controller, and the output is the resultant force. Its expression is as follows: ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2)) Wherein, e(k) is the heading angle error at the current moment, e(k-1) is the heading angle error at the previous moment, and e(k-2) is the heading angle error at the previous moment; ΔF=|F1|+|F2|, where F1 is the thrust allocated to the left main thruster (501), F2 is the thrust allocated to the right main thruster (502), and F1+F2=0.

7. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 6, characterized in that: The speed control of the multi-habitat unmanned platform includes speed maintenance and maneuvering speed change. Incremental PID control is used. The error between the actual speed and the expected value is used as the input of the PID controller, and the output is the resultant force. Its expression is as follows: ΔF(k)=K p (err(k)-err(k-1))+K i err(k)+K d (err(k)-2err(k-1)+err(k-2)) Wherein, e(k) is the speed error at the current moment, e(k-1) is the speed error at the previous moment, and e(k-2) is the speed error at the previous moment; ΔF=|F1|+|F2|, F1 is the thrust allocated to the left main thruster (501), and F2 is the thrust allocated to the right main thruster (502).

8. The active disturbance rejection control method for an amphibious unmanned platform with two-stage hybrid propulsion according to claim 7, characterized in that: In the horizontal rudder mode, the depth control of the multi-terrestrial unmanned platform is achieved through the vertical thrusters, horizontal rudders and main thrusters. An incremental PID control algorithm is used, with the depth error as the input of the PID controller and the horizontal rudder angle as the output. Combined with the line of sight navigation method, the 10m ahead of the multi-habitat unmanned platform's own X direction position is used as the desired X direction. d Direction position, the desired depth position is Z d , we get the displacement error equation: x e =x-x d ,z e =z-z d The desired pitch angle is: θ d <tan -1 ((zz d ) / (xx d ))。 9. A two-stage hybrid propulsion amphibious unmanned platform active disturbance rejection control system, characterized by: The system has a program module corresponding to the steps of any one of claims 1 to 8 above, and executes the steps in the above-mentioned two-stage hybrid propulsion amphibious unmanned platform self-disturbance rejection control method when running.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the self-disturbance rejection control method of the two-stage hybrid propulsion amphibious unmanned platform according to any one of claims 1 to 8 when called by a processor.

Images