Energy-saving track generation method of under-actuated AUV (Autonomous Underwater Vehicle)

Abstract

The invention discloses an energy-saving track generation method for an under-actuated AUV (autonomous underwater vehicle), and belongs to the technical field of autonomous underwater vehicles. The invention designs an AUV (Autonomous Underwater Vehicle) energy consumption optimal trajectory generator based on a direct collocation method aiming at the problem of three-dimensional energy-saving trajectory generation of an AUV under multiple constraints such as dynamic constraints. The method comprises the following steps of: firstly, performing complete analysis on an energy consumption generation mechanism of an AUV (Autonomous Underwater Vehicle) in executing a multi-path point tracking task according to related theories in combination with data obtained by a propeller open water experiment and CFD (Computational Fluid Dynamics) simulation; secondly, establishing a brand-new energy evaluation function, bringing the influence of time and attitude angles on energy consumption in the navigation process of the under-actuated AUV into the evaluation function, finally, adding AUV dynamic constraints to avoid the situation that a generated track cannot be tracked due to sharp corners or inflection points, and combining boundary constraints and path constraints of a system to obtain the energy consumption of the under-actuated AUV. An executable optimal energy trajectory is generated through offline trajectory planning based on a direct collocation method. The effectiveness and superiority of the method are verified through simulation experiments.

Description

Technical Field

[0001] The present invention belongs to the technical field of autonomous underwater vehicles, and in particular relates to an energy-saving trajectory generation method for an under-actuated AUV. Background Art

[0002] The endurance of an autonomous underwater vehicle (AUV) is one of its most important performance indicators, and it directly determines the operating cost of the vehicle.

[0003] Due to the limitation of the energy carried by AUV, 80% of the ocean on Earth has not been explored. How to improve the endurance of AUV when exploring the ocean has always been a difficult problem to be solved. The common method at present is to use the power of ocean currents or install a variable buoyancy system (VBS) to complete long-term exploration. However, this sacrifices the maneuverability of AUV, increases costs and lacks portability, so it is very meaningful to reduce the energy consumption of AUV from the perspective of control strategy.

[0004] In recent years, experts and scholars have conducted extensive research on the problem of AUV energy-saving trajectory generation. The technical route of most studies is to solve the optimal reference state of AUV by designing an efficient and feasible energy consumption cost function and minimizing the cost function. However, the premise of many current studies is to assume that the ocean current is slowly time-varying and the ocean environment information is known a priori, which is difficult to meet in actual engineering; in addition, different selections of energy consumption evaluation functions will also lead to differences in the final energy-saving effects. Simply taking the shortest navigation time or the minimum work done by the actuator as the minimum energy evaluation index is not rigorous for AUV. Therefore, it is extremely challenging and meaningful to refer to the efficient energy consumption evaluation function to generate a smooth energy-saving trajectory for AUV that meets the constraints of time, space and dynamic models at the planning level. Summary of the invention

[0005] The purpose of the present invention is to provide an energy-saving trajectory generation method for an underactuated AUV, which is characterized by optimal trajectory generation based on computational fluid dynamics simulation and direct matching method. Under the action of the energy-saving trajectory generator designed by the present invention, energy-optimal three-dimensional trajectory generation that meets the AUV dynamic constraints is achieved.

[0006] The present invention provides an energy-saving trajectory generation method for an underactuated AUV, comprising the following steps:

[0007] S110: Construct a five-degree-of-freedom mathematical model for the energy-saving trajectory generation task of underactuated AUV;

[0008] S120: analyzing the energy consumption of the underactuated AUV during navigation;

[0009] S130: Constructing the navigation resistance energy consumption model of underactuated AUV;

[0010] S140: List the optimal control problems to be solved in combination with the above steps and the specified path point information;

[0011] S150: Based on the direct collocation method, the continuous-time infinite-dimensional optimal control problem is discretized into a finite-dimensional nonlinear programming problem, and the optimal trajectory with the minimum energy consumption and the shortest time based on the original path points is generated through a solver.

[0012] Further, in S110, the five-degree-of-freedom model is:

[0013]

[0014] Where η = [x, y, z, θ, ψ] T is the position information of the underactuated AUV; ν=[u,v,w,q,r] T is the velocity information of the under-actuated AUV; α and γ are the attack angle and drift angle of the under-actuated AUV respectively; m ii , d ii , i∈[1,6] is the hydrodynamic coefficient; ρ is the seawater density; g is the gravitational acceleration; is the displacement volume; is the longitudinal metacenter height; control input τ=[τ u ,0,0,τ q ,τ r ] T .

[0015] Furthermore, in S120, the energy consumption of the underactuated AUV during navigation includes basic computing energy consumption, mission load energy consumption, energy consumption generated by the work of the underactuated AUV actuator during navigation, and energy consumption for overcoming navigation resistance; the navigation resistance energy consumption includes the work of the propeller to overcome the longitudinal resistance of the hull. And the increase in energy consumption due to the increase in angle of attack W d .

[0016] Furthermore, in S130, the propeller energy consumption for overcoming the longitudinal resistance of the hull is

[0017]

[0018] Where β is the power conversion ratio; u is the longitudinal speed; X uu is the longitudinal drag coefficient of the underactuated AUV; t0 is the start time of navigation; t f Stop time for navigation.

[0019] The energy consumption increase value W of the underactuated AUV due to the increase in the angle of attack is d for:

[0020]

[0021] Among them, ΔX is the increased longitudinal resistance of the underactuated AUV fitted by the least squares method; is the speed of the underactuated AUV;

[0022] The total energy consumption of the underactuated AUV during navigation is W auv :

[0023]

[0024] Among them, W R The energy consumed in overcoming resistance during navigation; W e For other energy consumption.

[0025] Furthermore, in S140, the energy optimal control problem is expressed as:

[0026]

[0027] stχ=f(t,χ(t),τ(t))

[0028]

[0029] where χ=[η,ν] T is the state of the AUV; t∈[t0,t f ] is the navigation time of the AUV; stχ=f(t,χ(t),τ(t)) is the dynamic constraint of the problem; is the boundary constraint of the problem, Upper limit for control input is the upper limit of the pitch angle, The upper limit of the termination time; is the path constraint of the problem; r COA is the switching radius; x(t j ),y(t j ),z(t j ) is the real-time position coordinate of AUV; WP = {WP i ∈R 3 ,i=0,…,N w} is the coordinate of the specified path point, WP0 is the initial position.

[0030] Furthermore, the S150 specifically includes the following steps:

[0031] S150.1: Divide the time domain of trajectory optimization into N equal intervals f segment, obtain N f+1 time-based collocation point, the time domain between each two time nodes is [t k-1 ,t k ],k=0,...,N f ;

[0032] S150.2: Rewrite the optimal control problem proposed in S140 into discrete form:

[0033]

[0034] st k+1 =f dt ·(t k ,η k ,v k ),η0=WP0

[0035]

[0036] Among them, f dt is the discrete-time vehicle kinematics obtained according to the time step Δt; WP0 is the initial position; is the upper bound of AUV speed and angular velocity; η mj is the mth discretized j state points; x mj ,y mj ,z mj is the real-time position of AUV; N w is the number of waypoint sequences.

[0037] S150.3: In [t0,t f ] divides the continuous time into segments N f segment, each subinterval is [t k ,t k+1 ],k=0,...,N f -1; collocation points for each subinterval

[0038]

[0039] Among them, h k =t k+1 -t k ;

[0040] S150.4: Use Simpson’s formula to calculate the subinterval t∈[t k ,t k+1 The state equation on Integrate to obtain the expected trajectory of the AUV in each sub-interval:

[0041]

[0042] Get the AUV at t∈[t0,t f Energy optimal trajectory η d (t):

[0043]

[0044] Among them, η d (t)=[x d ,y d ,z d ] T is bounded, η i (t), i = 0, 1, 2, ..., f-1 is the trajectory of each segment, and has bounded first-order and second-order derivatives with respect to time, and there exists a positive constant B0 that satisfies

[0045] Beneficial effects of the present invention:

[0046] Aiming at the problem of generating three-dimensional energy-saving trajectories for AUVs under multiple constraints such as dynamic constraints, the present invention designs an AUV energy-optimal trajectory generator based on the direct matching method. First, a complete analysis of the energy consumption mechanism of AUVs in performing multi-path point tracking tasks is conducted based on relevant theories combined with data obtained from open water experiments of thrusters and CFD simulations. Then, a new energy evaluation function is established, and the influence of time and attitude angles during the navigation of under-actuated AUVs is incorporated into the evaluation function. Finally, AUV dynamic constraints are added to avoid the occurrence of untrackable situations due to sharp corners or inflection points in the generated trajectory, and combined with the boundary constraints and path constraints of the system, an executable optimal energy trajectory is generated through DC-based offline trajectory planning. The effectiveness and superiority of this method are verified through simulation experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flow chart of a method for generating a three-dimensional energy-saving trajectory of an AUV with dynamic constraints provided in an embodiment of the present invention;

[0048] Figure 2 Schematic diagram of an open water experiment of an AUV thruster provided by an embodiment of the present invention;

[0049] Figure 3 1 is a schematic diagram of AUV "thrust-power" curve fitting provided by an embodiment of the present invention;

[0050] Figure 4 is a schematic diagram of computational fluid dynamics simulation of an AUV at zero angle of attack provided by an embodiment of the present invention;

[0051] Figure 5 is a schematic diagram of the “attack angle-resistance increase value” curve fitting of the AUV provided by an embodiment of the present invention;

[0052] Figure 6 It is a speed comparison diagram of the AUV energy-saving trajectory generator provided by an embodiment of the present invention;

[0053] Figure 7 It is a position and energy consumption comparison diagram of the AUV energy-saving trajectory generator provided by an embodiment of the present invention;

[0054] Figure 8 It is a position and energy consumption comparison diagram of the AUV energy-saving trajectory generator provided by an embodiment of the present invention when performing a multi-path point tracking task. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0056] like Figure 1 As shown, the present invention discloses an energy-saving trajectory generation method for an underactuated AUV, and the specific steps are as follows:

[0057] S110: Establish a five-degree-of-freedom mathematical model for under-actuated AUV trajectory generation; since the AUV has good symmetry, the influence of the roll angle on it is ignored, and a five-degree-of-freedom kinematic model is established.

[0058] S120: Analysis of the energy consumption mechanism of AUV.

[0059] The energy consumption of AUV during navigation mainly comes from the following parts:

[0060] (1) Basic computing energy consumption: generated by industrial computers and basic circuits;

[0061] (2) Mission payload energy consumption: generated when each sensor is operating;

[0062] (3) Energy consumption generated by the work done by the AUV actuators during navigation and energy consumption for overcoming navigation resistance;

[0063] It should be noted that the first two types of energy consumption are usually fixed or have small energy-saving potential under specified operating tasks. Therefore, this patent sets the power of the first two types of energy consumption to a fixed value, mainly to optimize the energy consumption generated during the navigation of the AUV.

[0064] S130: Design the navigation resistance energy consumption model of AUV;

[0065] The specific construction method of the AUV navigation resistance energy consumption model is as follows:

[0066] First, we refine the resistance of the AUV during navigation into the resistance along the longitudinal direction of the hull and the resistance increase caused by the angle of attack α of the hull. Overcoming the navigation resistance along the longitudinal direction of the hull is the effective thrust, expressed as X s =X |u∣u u 2 , and then, the empirical model P of propeller energy consumption can be used T =β(X s ) 1.5 Integral to calculate energy consumption.

[0067] The energy consumption caused by different attack angles α during the navigation of AUV is also different. The size of this value can be measured by computational fluid dynamics simulation software, and the increase value ΔX of the navigation resistance of AUV at different attack angles can be calculated through CFD numerical simulation, which provides a basis for the subsequent energy-saving trajectory generation.

[0068] By the least squares method, we know that given m groups of observation data (α i , ΔX i )(i=1,2,…,m), set the fitting model to ΔX=k q α 3 +k r α 2 , the parameter k is estimated by minimizing the following objective function q , k r :

[0069]

[0070] in, u, w are the longitudinal and vertical speeds of the AUV, respectively; β is the power conversion ratio of the propeller, which depends on the capability of each propeller and can be measured through relevant experiments; ΔX i is the observed value, is the fitted value, and S(a) is the residual sum of squares.

[0071] S140: Description of the optimal control problem; The specific construction method of the optimal control problem that the AUV needs to solve is as follows:

[0072] Based on the AUV mathematical model and the analysis and energy consumption model constructed in S110, S120, and S130, a multi-waypoint underwater exploration mission is considered. w waypoint sequence, WP = {WP i ∈R 3 ,i=0,…,N w}, given by the planning algorithm, where Contains the x, y and z positions of the ith waypoint, WP0 corresponds to the initial position of the vehicle. Assume that the vehicle is initialized with zero relative velocity and heading towards WP1. The energy optimal control problem is formulated as:

[0073]

[0074] stχ=f(t,χ(t),τ(t))

[0075]

[0076] where χ=[η,v] T is the state of the AUV; t∈[t0,t f ] is the navigation time of the AUV; stχ=f(t,χ(t),τ(t)) is the dynamic constraint of the problem; is the boundary constraint of the problem, Upper limit for control input is the upper limit of the pitch angle, The upper limit of the termination time; is the path constraint of the problem; r COA is the switching radius; x(t j ),y(t j ),z(t j ) is the real-time position coordinate of AUV; WP = {WP i ∈R 3 ,i=0,…,N w} is the coordinate of the specified path point, WP0 is the initial position.

[0077] S150: Generate a trajectory through a direct matching method, and verify its superiority through a simulation case; the specific construction method of the AUV energy-saving trajectory generation scheme based on the direct matching method is as follows:

[0078] In S150, we use offline trajectory planning based on the direct collocation method to solve the energy optimal control problem proposed in S140 and generate energy and time optimal trajectories.

[0079] First, the optimal control problem proposed in S140 is discretized according to the idea of ​​direct collocation method, and the original continuous-time infinite-dimensional optimal control problem (OCP) is converted into a finite-dimensional nonlinear programming problem (NLP). The time domain of trajectory optimization is divided into N equally spaced f segment, obtain N f +1 time-based collocation point (the time domain between each two time nodes is [t k-1 ,t k],k=0,...,N f ), so that the state variables and control variables of the system can be represented by the values ​​of specific collocation points. The following discrete variable node sequence is obtained:

[0080] t→t0…t k …t f ,

[0081]

[0082]

[0083] Then, the continuous optimal control problem proposed above is rewritten into a discrete form. In this way, solving the infinite-dimensional optimal control problem becomes solving a finite number of nonlinear programming problems, which greatly reduces the difficulty of solving.

[0084] Then, the optimal control problem is solved by an open source solver (IPOPT, etc.) to obtain a series of discrete states and control points. The Hermite polynomials are used for interpolation to express these result points in the form of state space equations. Finally, the Simpson formula is used to calculate the optimal control problem for each subinterval t∈[t k ,t k+1 The state equation on Integrate and obtain the energy optimal expected trajectory of the AUV.

[0085] Furthermore, simulation experiments also prove the superiority of this trajectory in energy saving.

[0086] Example 1

[0087] The present invention discloses a method for generating an energy-saving trajectory of an underactuated AUV, comprising the following steps:

[0088] S110: Establish a five-degree-of-freedom mathematical model for underactuated AUV trajectory generation; since the AUV has good symmetry, the effect of the roll angle on it is ignored, and its five-degree-of-freedom kinematic model is:

[0089]

[0090] The kinetic equation is:

[0091]

[0092] Where η = [x, y, z, θ, ψ] T represents the position vector of the AUV in the Earth-fixed inertial system, ν = [u, v, w, q, r] T represents the velocity vector of the AUV in the Earth-fixed inertial system, α and γ are the attack angle and drift angle of the AUV respectively, and m ii ,dii , i∈[1,6] is the measurable hydrodynamic coefficient, ρ is the seawater density, g is the gravitational acceleration, is the displacement volume, is the longitudinal metacenter height, the control input τ=[τ u ,0,0,τ q , τ r ] T .

[0093] S120, analysis of the energy consumption mechanism during AUV navigation.

[0094] The energy consumption of AUV during navigation mainly comes from the following parts:

[0095] (1) Basic computing energy consumption: generated by industrial computers and basic circuits;

[0096] (2) Mission payload energy consumption: generated when each sensor is operating;

[0097] (3) Energy consumption generated by the work done by the AUV actuators during navigation and energy consumption for overcoming navigation resistance;

[0098] It should be noted that the first three types of energy consumption are usually fixed or have small energy saving potential under a specified task. Therefore, we consider reducing the navigation energy consumption of AUV during task operation at the control strategy level.

[0099] See also Figure 2 As shown, S130 is performed, AUV "thrust-power" curve fitting scheme;

[0100] This paper uses a certain type of DC motor driven duct propeller for simulation analysis. Its power conversion ratio β = 0.4507, which is obtained from the "thrust-power" experimental data measured by the propeller in the open water test of the circulating water tank, and is calculated based on the empirical model P T =β(X s ) 1.5 Fitted, where X s =X |u∣u u 2 ,like Figure 3 shown.

[0101] Furthermore, the energy consumption caused by different attack angles α during the navigation of the AUV is also different; the present invention will give an example in combination with an actual AUV boat type, such as Figure 4 , 5 As shown, the resistance of the AUV in different navigation postures of this embodiment is calculated through CFD numerical simulation, providing a basis for energy-saving control.

[0102] Using the nonlinear least squares method to solve the problem, we can obtain the resistance increase curves at different attack angles relative to the 0° attack angle, as well as the approximate relationship between the attack angle α and the increase in longitudinal resistance ΔX:

[0103] ΔX=-0.002315α 3 +0.05841α 2

[0104] In the formula, u and w are the longitudinal and vertical velocities of the AUV, respectively;

[0105] The data obtained through CFD simulation clearly shows that the angle of attack of the AUV is positively correlated with the resistance curve. Therefore, if we want to reduce the energy consumption caused by navigation resistance, we need to make the angle of attack of the AUV as small as possible during navigation.

[0106] In this paper, the power of other sensors and basic circuits except the thruster is set to a fixed value P e , this part of energy consumption can be obtained by integrating the navigation time. In this way, the comprehensive energy consumption of AUV during navigation can be obtained as:

[0107]

[0108] Where W auv Represents the total energy consumption of the AUV system when performing tasks, W R is the energy consumption caused by sailing resistance during sailing, W e represents other energy consumption, t0, t f is the sailing start and stop time, Represents the speed of the AUV.

[0109] S140, the specific construction method of the optimal control problem that the AUV needs to solve is as follows:

[0110] Based on the AUV mathematical model and the analysis and energy consumption model constructed in S110, S120, and S130, a multi-waypoint underwater exploration mission is considered. w waypoint sequence, WP = {WP i ∈R 3 ,i=0,…,N w}, given by the planning algorithm, where Contains the x, y and z positions of the ith waypoint, WP0 corresponds to the initial position of the vehicle. Assume that the vehicle is initialized with zero relative velocity and heading towards WP1. The energy optimal control problem is formulated as:

[0111]

[0112] Where χ=[η,ν]T is the state of the AUV, is the upper limit of the pitch angle and the termination time, and the remaining formulas are the AUV dynamics constraint, boundary constraint, and path constraint.

[0113] S150, in this section we use offline trajectory planning based on the direct collocation method to solve the energy optimal control problem proposed in S140 and generate energy and time optimal trajectories.

[0114] First, the optimal control problem proposed in S140 is discretized according to the idea of ​​direct collocation method, and the original continuous-time infinite-dimensional optimal control problem (OCP) is converted into a finite-dimensional nonlinear programming problem (NLP). The time domain of trajectory optimization is divided into N equally spaced f segment, obtain N f +1 time-based collocation point (the time domain between each two time nodes is [t k-1 ,t k ],k=0,...,N f ), so that the state variables and control variables of the system can be represented by the values ​​of specific collocation points. The following discrete variable node sequence is obtained:

[0115] t→t0…t k …t f ,

[0116]

[0117]

[0118] Then, considering the kinematic model constraints of the AUV system, the terminal state is fixed and the terminal time t f The variable boundary constraints and the path constraints through fixed path points rewrite the optimal control problem proposed above into a discrete form as shown below:

[0119]

[0120] st k+1 =f dt ·(t k ,η k ,v k ),η0=WP0

[0121]

[0122] In the formula, f dt · is the discrete-time vehicle kinematics obtained in time steps Δt.

[0123] Furthermore, by solving this optimal control problem through an open source solver (IPOPT, etc.), the optimized control variables and state variables of the AUV that performs the target task under multiple constraints can be obtained.

[0124] Then, in [t0,t f ] divide the continuous time into N f segment, each subinterval is [t k ,t k+1 ],k=0,1,...,N f -1. Let h k =t k+1 -t k , Then s∈[0,1]. In this subinterval, each state variable can be expressed by a cubic Hermite polynomial:

[0125] η(t)=c k,0 +c k,1 s+c k,2 s 2 +c k,3 s 3

[0126] Among them, c k,i is the coefficient to be sought.

[0127] According to the boundary condition η k =η(0),η k+1` =η(1), It can be concluded that:

[0128]

[0129] Selecting the midpoint of each subinterval as the collocation point yields:

[0130]

[0131] Using Simpson's formula for the subinterval t∈[t k ,t k+1 The state equation on By integrating, we can obtain the expected trajectory of the AUV in each sub-interval:

[0132]

[0133] In summary, we can get that AUV is f Energy optimal trajectory η d (t):

[0134]

[0135] Among them, ηd (t) = [x d ,y d ,z d ] T is bounded, η i (t), i = 0, 1, 2, ..., f-1 is the trajectory of each segment. And it has bounded first-order and second-order derivatives with respect to time, and there exists a positive constant B0 that satisfies

[0136] The feedforward trajectory generated by the direct matching method is under ideal conditions (model parameters are known and fixed, there is no external disturbance and actuator saturation). The trajectory generated by this method is an open-loop feedforward trajectory, which does not take into account the complex and changeable marine environment. The control input obtained cannot directly act on the AUV system, but the generated energy-saving trajectory η d (t) The desired trajectory designed as a subsequent controller can still reduce the energy consumption of the AUV during the mission.

[0137] In order to verify the advancedness of the proposed energy consumption model, the multi-waypoint tracking task problem proposed in S150 is considered. Trajectory generation is performed by considering only two waypoints at a time. In this way, the performance of different energy consumption evaluation functions can be more intuitively compared. We conduct three sets of comparative experiments on minimizing thrust, minimizing time, and the method proposed in this paper:

[0138] #1)

[0139] #2)minJ2=t f -t0

[0140] #3)

[0141] Among them, the propeller power conversion ratio is measured by the previous water tank experiment, β = 0.4507, and the remaining energy consumption generated during navigation is P e =0.5W.

[0142] We can assume that the waypoints are known a priori, and select the set waypoints as: WP1 = (0, 0, 0), WP2 = (50, 10, 30), WP3 = (0, 30, 0) to set the initial position and attitude of the AUV to η0 = [x, y, z, ψ, θ] T =[0,0,0,0,0] T , the initial velocity is v0 = [u,v,w,q,r] T =[0,0,0,0,0] T , the constraint parameters are chosen as r COA =2m. Number of matching points N f The number of selections is 200.

[0143] The energy consumption and corresponding AUV trajectory generated by different evaluation functions under the same constraints are shown in Figure 2. Figure 7 As shown, the corresponding speed is Figure 6 It can be seen that the three sets of energy consumption evaluation functions can access the waypoints in a satisfactory state under the same constraints. On this basis, compared with the energy consumption evaluation functions used in #1 and #2, the method proposed in this paper has a better performance in terms of propulsion energy consumption W. thrust , flight time t f , and the average angle of attack The three aspects of change rate have better comprehensive performance, which improves the performance by nearly 48.13% compared with energy consumption evaluation function #1.

[0144] Furthermore, on this basis, we select the problem of multi-path point tracking task and consider setting the path points:

[0145] WP1=(0,0,0),WP2=(50,0,30),WP3=(50,10,30),

[0146] WP4=(0,10,0),WP5=(0,20,0),WP6=(50,20,30),WP7=(50,30,30),WP8=(0,30,0)

[0147] Assume that there is a steady ocean current U in the environment c =1.2m / s,ψ c =0.5rad,θ c =0, set the maximum speed V that the AUV can reach max =2m / s, the switching radius r is set to 2m, and the energy-saving trajectory generator proposed in the present invention is compared with other similar energy-saving methods combined with the same feedback controller. The comparison results are as follows Figure 8 As shown, it can be clearly seen that the solution proposed by the present invention improves the energy utilization efficiency by 5.7% compared with the method obtained by optimizing the speed, and improves the energy utilization efficiency by 32.9% compared with the traditional LOS guidance method. This shows that the trajectory generator proposed by the present invention has a higher energy-saving effect, that is, it can better handle the energy-saving trajectory generation problem of underactuated AUV under multi-objective constraints.

[0148] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should fall within the scope of the patent coverage of the present invention.

Claims

1. A method for generating energy-saving trajectories for an underactuated AUV, characterized by: The following steps are involved: S110: Construct a five-degree-of-freedom mathematical model for the energy-saving trajectory generation task of underactuated AUV; S120: analyzing the energy consumption of the underactuated AUV during navigation; S130: Constructing the navigation resistance energy consumption model of underactuated AUV; S140: combining the above steps and the specified path point information, listing the energy optimal control problems to be solved; S150: Based on the direct collocation method, the continuous-time infinite-dimensional energy optimal control problem is discretized into a finite-dimensional nonlinear programming problem, and the optimal trajectory with the minimum energy consumption and the shortest time based on the original path points is generated through a solver.

2. The method for generating energy-saving trajectories for underactuated AUVs according to claim 1, characterized in that: In S110, the five-degree-of-freedom model is: Where η = [x, y, z, θ, ψ] T is the position information of the underactuated AUV; ν=[u,v,w,q,r] T is the velocity information of the under-actuated AUV; α and γ are the attack angle and drift angle of the under-actuated AUV respectively; m ii , d ii , i∈[1,6] is the hydrodynamic coefficient; ρ is the seawater density; g is the gravitational acceleration; is the displacement volume; is the longitudinal metacenter height; control input τ=[τ u ,0,0,τ q ,τ r ] T .

3. The energy-saving trajectory generation method for an underactuated AUV according to claim 2, characterized in that: In S120, the energy consumption of the underactuated AUV during navigation includes basic computing energy consumption, mission load energy consumption, energy consumption generated by the work of the underactuated AUV actuator during navigation, and energy consumption for overcoming navigation resistance; the navigation resistance energy consumption includes the work of the propeller to overcome the longitudinal resistance of the hull. And the increase in energy consumption due to the increase in angle of attack W d .

4. The energy-saving trajectory generation method for an underactuated AUV according to claim 3, characterized in that: In S130, the propeller energy consumption to overcome the longitudinal resistance of the hull is Where β is the power conversion ratio; u is the longitudinal speed; X uu is the longitudinal drag coefficient of the underactuated AUV; t0 is the start time of navigation; t f Stop time for navigation. The energy consumption increase value W of the underactuated AUV due to the increase in the angle of attack is d for: Among them, ΔX is the increased longitudinal resistance of the underactuated AUV fitted by the least squares method; is the speed of the underactuated AUV; The total energy consumption of the underactuated AUV during navigation is W auv : Among them, W R The energy consumed in overcoming resistance during navigation; W e For other energy consumption.

5. The energy-saving trajectory generation method for an underactuated AUV according to claim 3, characterized in that: In S140, the energy optimal control problem is expressed as: stχ=f(t,χ(t),τ(t)) where χ=[η,ν] T is the state of the AUV; t∈[t0,t f ] is the navigation time of the AUV; stχ=f(t,χ(t),τ(t)) is the dynamic constraint of the problem; is the boundary constraint of the problem, is the upper limit of the control input, is the upper limit of the pitch angle, The upper limit of the termination time; is the path constraint of the problem; r COA is the switching radius; x(t j ),y(t j ),z(t j ) is the real-time position coordinate of AUV; WP = {WP i ∈R 3 ,i=0,L,N w } is the coordinate of the specified path point, WP i =(WP i x ,WP i y ,WP i z ), WP0 is the initial position.

6. The energy-saving trajectory generation method for an underactuated AUV according to claim 4, characterized in that: The S150 specifically includes the following steps: S150.1: Divide the time domain of trajectory optimization into N equal intervals f segment, obtain N f +1 time-based collocation point, the time domain between each two time nodes is [t k-1 ,t k ],k=0,K,N f ; S150.2: Rewrite the optimal control problem proposed in S140 into discrete form; st k+1 =f dt ·(t k ,or k ,v k ),η0=WP0 Among them, f dt is the discrete-time vehicle kinematics obtained according to the time step Δt; WP0 is the initial position; v is the upper bound of the AUV velocity and angular velocity; η mj is the mth discretized j state points; x mj ,y mj ,z mj is the real-time position of AUV; N w is the number of waypoint sequences. S150.3: In [t0,t f ] divides the continuous time into segments N f segment, each subinterval is [t k ,t k+1 ],k=0,K,N f -1; collocation points for each subinterval Among them, h k =t k+1 -t k ; S150.4: Use Simpson’s formula to calculate the subinterval t∈[t k ,t k+1 The state equation on Integrate to obtain the expected trajectory of the AUV in each sub-interval: Get the AUV at t∈[t0,t f Energy optimal trajectory η d (t): Among them, η d (t) = [x d ,y d ,z d ] T is bounded, η i (t), i = 0, 1, 2, L, f-1 is the trajectory of each segment, and has bounded first-order and second-order derivatives with respect to time, and there exists a positive constant B0 that satisfies

Images