Underwater thruster thrust distribution method based on multi-motor cooperation

Abstract

The invention discloses an underwater propeller thrust distribution method based on multi-motor coordination, which comprises the following steps: inputting an expected attitude given by an AUV (Autonomous Underwater Vehicle) multi-motor coordination propelling system and an actual feedback attitude of the AUV into an AUV attitude controller to obtain a control force signal; the thrust distributor decomposes the control force signal in real time, converts the control force signal into a rotating speed signal and transmits the rotating speed signal to the AUV multi-motor coordinated propulsion system; establishing a thrust control distribution problem of the AUV propeller; and solving a thrust control distribution problem by adopting a variable target AUV thrust distribution solving algorithm, and outputting a thruster thrust distribution result. According to the method, the overall navigation performance of the AUV in a complex underwater environment can be improved, particularly, the optimal balance between energy consumption optimization and thrust distribution precision can be sought, the optimization target can be dynamically adjusted according to the real-time navigation requirement, and the optimal compromise of energy consumption and precision is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of underwater robot control, relates to the thrust distribution technology of underwater thrusters, and particularly relates to a thrust distribution method for underwater thrusters based on multi-motor coordination. Background Art

[0002] With the continuous increase in the global intensity of ocean resource development, the demand for high-performance underwater vehicles is increasing day by day. These vehicles need to have stronger propulsion capabilities, higher stability, and longer endurance to cope with complex ocean environments. The underwater multi-motor propulsion technology has been developed to meet these needs. However, the underwater operating environment of underwater vehicles is relatively complex, and the thrust generated by the current AUV (Autonomous Underwater Vehicle) thrust distribution method may not reach the thrust expected by the control system, which may also lead to insufficient power supply of the AUV and the failure of underwater operation tasks.

[0003] The existing AUV thruster thrust distribution methods are mainly the pseudo-inverse method and the quadratic programming method. The pseudo-inverse method is mainly applicable to low underwater disturbances and can quickly obtain the thrust distribution scheme. However, when the actuator control amount is saturated, it may not be able to automatically adjust the remaining control amounts to compensate for the deficiency, resulting in the actual control force not reaching the expected value. Although the quadratic programming method can effectively handle constraint conditions such as thruster saturation, its calculation delay is relatively significant, which may affect the real-time dynamic performance of the AUV. Summary of the Invention

[0004] Object of the Invention: In order to overcome the deficiencies in the prior art, the present invention provides a thrust distribution method for underwater thrusters based on multi-motor coordination, which can improve the overall navigation performance of the AUV in complex underwater environments, especially can find the best balance between energy consumption optimization and thrust distribution accuracy, and can dynamically adjust the optimization target according to real-time navigation requirements to achieve the optimal compromise between energy consumption and accuracy.

[0005] Technical Solution: To achieve the above object, the present invention provides a thrust distribution method for underwater thrusters based on multi-motor coordination, including the following steps:

[0006] S1: Input the expected attitude given by the AUV multi-motor coordinated propulsion system and the actual feedback attitude of the AUV into the AUV attitude controller to obtain the control force signal;

[0007] S2: The thrust distributor decomposes the control force signal in real time and converts it into a rotational speed signal to be transmitted to the AUV multi-motor coordinated propulsion system;

[0008] S3: Establish the thrust control distribution problem of the AUV thruster;

[0009] S4: Solve the thrust control distribution problem by using a variable - target AUV thrust distribution solution algorithm, and output the thruster thrust distribution result.

[0010] Further, in step S1, the AUV multi - motor coordinated propulsion system is equipped with multiple thrusters, and their layout and design are as follows:

[0011] The number of thrusters is 7, namely thruster No. 1 - No. 7. Thruster No. 1 and No. 2 are paddle - rudder combined thrusters of the AUV. They are arranged along the x - axis direction, and the dynamic control of the thrust direction of a single thruster of the AUV is realized by adjusting the rudder angle. Their control angles α1, α2 ∈ [-90°, 90°], and the clockwise direction is positive; Thruster No. 3 and No. 4 are fixed - angle side - thrusters, which are symmetrically distributed with respect to the x - axis and the included angle is taken as β1 = β2 = 45°, providing lateral thrust compensation for the AUV; Thrusters No. 1 to No. 4 are all horizontal thrusters, providing X - Y - N degree - of - freedom control forces and torques for the AUV propulsion system; Thrusters No. 5 to No. 7 are vertical thrusters and are arranged in a triangle along the z - axis. The combined action of these three vertical thrusters provides Z - M - K degree - of - freedom control forces and torques for the AUV.

[0012] Further, in the AUV multi - motor coordinated propulsion system, a mathematical model is established for a single thruster:

[0013] n i is the propeller speed, where the power P of the AUV thruster L and the torque T L The relationship is expressed as:

[0014] P Li = 2πn i T Li (1)

[0015] Determine the relationship between the propeller speed and the thrust it generates, and establish the expected thrust and torque distribution formula:

[0016]

[0017] In the formula, D i is the propeller diameter of each thruster, ρ is the seawater density, K T and K Q are torque - related coefficients;

[0018] Combining formula (1) and (2) gives:

[0019]

[0020] Further, in the AUV multi - motor coordinated propulsion system, establish the relationship between the thrust and torque of the AUV six - degree - of - freedom motion output and the given thrust of the thruster, and use this relationship as the equality constraint condition in the AUV thrust distribution. The specific process includes:

[0021] A1: First, start the calculation of the combined thrust and moment of the horizontal thrusters. Define the thrust generated by each thruster as F i , according to the actual power layout of thrusters 1 to 4 in the horizontal direction, where thrusters 1 and 2 are symmetrically distributed about the X-axis to generate and balance the rotational moment; d xi and d yi respectively represent the longitudinal and lateral positions of each thruster on the AUV. Calculate the magnitudes of the combined thrust and combined moment of the six-degree-of-freedom motion generated by the four horizontally arranged thrusters, expressed as:

[0022]

[0023] In the above formula represents the thrust vector of the single-target output of the seven thrusters arranged on the AUV, is the resultant force and resultant moment received by the AUV in the horizontal direction;

[0024] A2: Conduct the calculation of the combined thrust and moment of the vertical thrusters. To facilitate the calculation, consider the case where the vertically arranged thrusters are coplanar, and do not consider the interference of the horizontal thrusters on them. Here, arrange the vertical thrusters on the AUV in a triangular shape along the z-axis to meet the power requirements; calculate the magnitudes of the combined thrust and moment of the six-degree-of-freedom motion generated by the three vertically arranged thrusters, expressed as:

[0025]

[0026] In the above formula is the resultant force and resultant moment received by the AUV in the vertical direction;

[0027] A3: Finally, calculate the combined thrust and resultant moment of the thrusters at seven different positions on the AUV propulsion system. Use the vector sum to obtain the combined thrust and resultant moment generated by the four horizontal thrusters and the three vertical thrusters, expressed as:

[0028] In the above formula [F X F Y F Z T K T M T N is the resultant force and resultant moment received by the AUV;

[0029] Finally, the relationship between the thrust and moment output of the AUV's six-degree-of-freedom motion and the given thrust F i of the thrusters is:

[0030]

[0031] Equation (6) is the equality constraint condition in the AUV thrust allocation. τ is the AUV output thrust and thrust moment vector, and B(α,β) is the AUV thruster vector layout matrix.

[0032] Furthermore, based on the equality constraint condition in the AUV thrust allocation in step S3, the thrust control allocation problem of the AUV thruster is established as follows:

[0033] B1: The AUV thrust allocation aims to minimize the thruster energy consumption. According to the relationship between the power and thrust of the AUV thruster, Equation (3) is further simplified to obtain:

[0034] P L = χF Ti 1.5 (7)

[0035] where is a fixed constant, treated equivalently to ; for the convenience of subsequent calculations, here the relationship between the AUV thruster power and thrust is replaced by a quadratic form as Here, taking the minimum power consumption of the AUV thruster as the optimization goal, the AUV thrust allocation objective function is established as:

[0036] minf = F Ti T MF Ti (8)

[0037] where f is the power consumption of the AUV thruster; F T is the thrust vector of the AUV thruster, and M is a positive definite weighted coefficient matrix used to determine the weight of the AUV thruster energy consumption in the objective function;

[0038] B2: Continue to optimize the AUV thrust allocation error. Here, when the desired forces and torques of each AUV thruster change greatly, the thruster may not be able to adjust the rotation speed and azimuth angle in time to match the current required τ. To solve this problem, a slack variable s is introduced to ensure that the AUV thrust allocation problem has a feasible solution, expressed as:

[0039]

[0040] where s is the difference between the control command generated by the AUV controller and the actual thrust after thrust allocation. If the difference is too large, it cannot be guaranteed that the AUV sails on the expected trajectory or stays at the desired position. An AUV thrust allocation error function term is introduced:

[0041] J s = s T Qs (10)

[0042] Among them, Q is a positive definite weighting matrix used to ensure that the AUV thrust allocation error is controlled to zero;

[0043] B3: The AUV thruster thrust allocation equality constraint is expressed as The inequality constraint conditions consider the maximum thrust limit of the single-objective output of each thruster, the thrust change rate of the AUV thruster in a short time, and the thruster rudder angle change range limit, and are expressed as:

[0044]

[0045] Among them, F Ti is the magnitude of the thrust output by the i-th thruster on the AUV, F Timax and F Timin are the maximum and minimum thrusts output by the i-th thruster respectively; ΔF Ti is the thrust change rate of the AUV thruster in a short time, ΔF Timax and ΔF Timin are the maximum and minimum values of the change in the thrust of the i-th thruster in a short time respectively; α i is the magnitude of the rudder angle of the i-th thruster on the AUV, α imax and α imin are the maximum and minimum rudder angles output by the i-th thruster respectively;

[0046] Finally, the objective function of the AUV thrust allocation problem is obtained as:

[0047]

[0048] The above inequality constraints include the thruster thrust change range constraint and the thruster saturation constraint, and the equality constraint realizes the controllability of the AUV attitude motion and the thrust error constraint.

[0049] Furthermore, the solution method for solving the thrust control allocation problem by using the variable-objective AUV thrust allocation solution algorithm in the step S4 includes:

[0050] Case 1: When the AUV thruster thrust allocation is not saturated or not constrained, with the goal of minimizing energy consumption, the method of the solution space of the equations is used to obtain the optimal solution of the thrust allocation problem;

[0051] Case 2: When thrust saturation occurs, it changes from the method of Case 1 to the goal of minimizing error, and the optimal solution of the thrust allocation problem is solved by the active set quadratic programming method;

[0052] Case 3: When a thruster fails, directly with the goal of minimizing the error of the AUV thruster thrust, the optimal solution of the thrust allocation problem is solved by the active set quadratic programming method.

[0053] Furthermore, when it comes to Case 1 in step S4, the specific solution process of the thrust control allocation problem is as follows:

[0054] C1: Adopt the AUV thruster thrust optimization allocation strategy based on minimum energy consumption. According to the AUV thrust space layout matrix B(α,β), list the homogeneous equation system related to AUV thrust allocation B i (α,β)F Ti =τ i , this equation system consists of the general solution FT0 and the particular solution FT*. The general solution is that the kernel of the AUV thrust space layout matrix B(α,β) is F T0 =Ker(B(α,β)), and the particular solution is directly obtained by the pseudoinverse method as F T * =B(α,β) + τ, and the solution of the equation system is expressed as the sum of the general solution and the particular solution:

[0055] F Ti =χ i Ker(B i (α,β))+B i (α,β) + τ (13)

[0056] where χ i is the set of valid solutions in the AUV thrust allocation problem, and thus find the minimum energy consumption solution; write the inequality constraint of the general solution of the equation system according to formula (11):

[0057] F Timin -B i (α,β) + τ≤χKer(B i (α,β))≤F Timax +B i (α,β) + τ (14)

[0058] When the value of χ i exists, it indicates that there is no saturation problem in thrust allocation, and the valid solution in the equation system is obtained by the pseudoinverse method. Otherwise, if it does not exist, the quadratic programming method is used to solve the equation system;

[0059] C2: According to the quadratic form model of formula (8), with the minimum energy consumption of AUV thrust allocation as the control optimization goal, express the AUV thrust allocation objective function as:

[0060]

[0061] Take the inner product of the general solution and the particular solution in the equation of formula (15) to obtain:

[0062] Ker(B i (α,β))T B i (α, β) + τ = (B i (α, β)Ker(B i (α, β))) T (B(α, β)B(α, β) -1 ) T τ(16)

[0063] Since the kernel of the general solution is the zero vector, we get B i (α, β)Ker(B i (α, β)) = 0, and further calculation gives:

[0064] Ker(B i (α, β)) T B i (α, β) + τ = 0 (17)

[0065] Under this condition, the minimum energy consumption problem is equivalently described as finding a solution that minimizes the energy consumption while satisfying the given conditions. Its optimal energy consumption problem is expressed as:

[0066]

[0067] Equivalent the above minimum energy consumption problem to the problem of the value of χ i When χ i ≠ 0, the minimum energy consumption solution of AUV thrust allocation is expressed as:

[0068]

[0069] When χ i = 0, the minimum energy consumption solution of AUV thrust allocation is solved by the pseudoinverse method and expressed as:

[0070]

[0071] The above is the particular solution of the system of equations space method, and it can only be solved when the AUV thruster thrust allocation is unsaturated or unconstrained.

[0072] Furthermore, in the case of situation two in step S4, the specific solution process of the thrust control allocation problem is as follows:

[0073] Based on minimizing the AUV thruster thrust error for optimal allocation, mainly aiming at the situation that the system of equations space method cannot solve the AUV thruster thrust saturation situation. Here, when the thrust is saturated, the active set in the quadratic programming method is used to find the optimal solution. The AUV thrust optimal allocation problem is expressed by quadratic programming as:

[0074]

[0075] In the formula, in the control objective function, W is an n-order symmetric matrix determined by the quadratic coefficient in the objective function; E and G are respectively subsets of equality constraints and inequality constraints, and A, B, C, and s are also n-order vectors;

[0076] Then, it is rewritten into an incremental form through the Taylor formula as:

[0077]

[0078] The above optimal and effective particular solution of AUV thrust allocation needs to satisfy the KT (Kuhn Tucker) condition:

[0079]

[0080] Among them, is the search gradient of the Lagrangian function in the AUV thrust allocation optimization problem, and h(s * ) ≥ 0 represents the inequality constraint on the optimal allocation of AUV thrust, and it is a necessary and sufficient condition for s * to be an effective solution of the system of equations; when χ i ≥ 0 is the dual feasibility, we get And when χ i < 0, it means that the search gradient descent direction of the control objective function is consistent with the gradient direction of the constraint boundary. Therefore, the obtained solution is not the optimal solution, that is, it does not satisfy the dual effectiveness, so s * is not the optimal solution of the thrust allocation problem; when the optimal solution is inside the control domain, that is, h(s * ) ≥ 0, it is equivalent to an unconstrained optimization problem. At this time, χ i = 0; on the contrary, if the optimal solution in the thrust allocation problem is located on the boundary of the control domain, that is, h(s * ) = 0, then the optimal solution problem must satisfy that the product of χi and h(s * ) is zero, that is, the complementary slackness;

[0081] The AUV thrust optimization problem is transformed into an active set algorithm in quadratic programming for optimization and solution. The active set method includes initializing to provide an initial feasible solution s0 and an initial working set G0, taking the optimal solution in the thrust allocation problem as the initial solution of the optimization problem, and setting the initial set as an empty set; first, find the search direction δ k = s k - s k-1 and the corresponding Lagrange multiplier η i , and the iteration step size is defaulted to O k= 1; Then, transform the inequality constraints of the working set G0 into equality constraints that only restrict the boundary conditions, solve for the global minimum solution s' in its sub-problem, and set the corresponding Lagrange multiplier η i ; If |s' - s k | < ε, where ε is an infinitesimal greater than zero, and χ i ≥ 0, then the optimal solution to the AUV thrust allocation problem is obtained. Update the corresponding active set according to the sign of the Lagrange multiplier χ i , and remove the ineffective constraint opportunities from G k * ; If the new feasible solution is not within the feasible region, select an appropriate step size in the current search direction to bring it back into the feasible region. The process is expressed by the formula:

[0082]

[0083] If |s' - s k | > ε, then s' is a solution on the feasible region of the thrust allocation problem, and s k+1 = s'. Through the above formula iteration update, if s' is not a solution on the feasible region, the following assumptions are made according to formula (1.25):

[0084]

[0085] Its iteration rule is:

[0086]

[0087] Finally, based on the AUV thruster power layout and mutual constraint parameters, combined with the minimum AUV thruster thrust allocation optimization method, the optimization model is obtained as:

[0088]

[0089] Furthermore, when a thruster fails in step S4, that is, in case three, the originally possible fully actuated thrust configuration may change to an underactuated state. In this case, the thrust allocation problem may become unsolvable. To address this situation, the active set method can be directly used to solve the optimal solution with the goal of minimizing the AUV thruster thrust error:

[0090] Remove the inequality constraints of the AUV thrusters in the following formula:

[0091] F Timin ≤ F Ti ≤ F Timax (28)

[0092] And add the equality constraints of the corresponding failed thrusters:

[0093] F Ti = 0 (29)

[0094] In the present invention, a variable target thrust optimization distribution method is proposed to achieve the scenario - adaptive balance between energy consumption and accuracy, so as to realize the accurate distribution of the thrust of the AUV thrusters. By designing an accurate AUV thruster model and determining the AUV thrust distribution process, the power source for the AUV's underwater operation is ensured. Then, the power layout of the AUV propulsion system and the problem of thruster thrust control distribution are determined. Finally, a variable target AUV thruster thrust distribution method is proposed. When there is no thrust saturation in the AUV thruster thrust distribution, a thrust distribution with minimized energy consumption and good real - time performance is adopted; when thrust saturation occurs, the strategy switches to minimizing the error distribution. This provides a new solution for the application of the AUV propulsion system in the coordinated operation of multiple propulsion motors.

[0095] Advantageous effects: Compared with the prior art, in the present invention, by constructing an accurate thruster model and determining the AUV thrust distribution process, and further building the power layout of the AUV propulsion system to facilitate the AUV to complete six - degree - of - freedom spatial motion. Based on the AUV power layout, the basic problems of AUV thruster thrust distribution are analyzed, and a thrust distribution algorithm composed of variable targets combining the minimum energy consumption and minimum error of the AUV is proposed. The present invention adopts a thrust distribution algorithm composed of variable targets of minimum energy consumption and minimum error, which combines the advantages of the pseudo - inverse method and the quadratic programming method, and can flexibly adjust the optimization method under different working conditions. This method aims to dynamically adjust the optimization target according to the real - time navigation requirements to achieve the optimal compromise between energy consumption and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] Figure 1 It is a block diagram of the AUV thrust distribution system in an embodiment of the present invention;

[0097] Figure 2 It is a power layout diagram of multiple propulsion motor thrusters in an embodiment of the present invention;

[0098] Figure 3 It is a working process diagram of variable target thrust optimization distribution in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0099] The following further clarifies the present invention in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. After reading the present invention, various equivalent forms of modification by those skilled in the art fall within the scope defined by the appended claims of this application.

[0100] The present invention provides an underwater thruster thrust distribution method based on multi - motor cooperation. Referring to Figure 1 , it includes the following steps:

[0101] S1: The desired attitude given by the AUV multi-motor coordinated propulsion system and the actual feedback attitude of the AUV are input into the AUV attitude controller to obtain the control force signal;

[0102] S2: The thrust distributor decomposes the control force signal in real time and converts it into a rotational speed signal to be transmitted to the AUV multi-motor coordinated propulsion system;

[0103] S3: Establish the thrust control distribution problem of the AUV thrusters;

[0104] S4: Use the variable-target AUV thrust distribution solution algorithm to solve the thrust control distribution problem and output the thruster thrust distribution result.

[0105] In step S1, the AUV multi-motor coordinated propulsion system is equipped with multiple thrusters, and its layout and design are as follows:

[0106] As Figure 2 shown, the number of thrusters is 7, namely thruster No. 1 to No. 7. Thruster No. 1 and No. 2 are propeller-rudder combined thrusters of the AUV. They are arranged along the x-axis direction, and the dynamic control of the thrust direction of a single AUV thruster is realized by adjusting the rudder angle. Its control angles α1, α2 ∈ [-90°, 90°] and the clockwise direction is positive; Thruster No. 3 and No. 4 are fixed-angle side thrusters, symmetrically distributed with the x-axis and the included angle is taken as β1 = β2 = 45°, providing lateral thrust compensation for the AUV; Thruster No. 1 to No. 4 are all horizontal thrusters, providing X-Y-N degree-of-freedom control forces and torques for the AUV propulsion system; Thruster No. 5 to No. 7 are vertical thrusters and are arranged in a triangle along the z-axis. The combined action of these three vertical thrusters provides Z-M-K degree-of-freedom control forces and torques for the AUV.

[0107] Mathematical modeling of a single thruster in the AUV multi-motor coordinated propulsion system:

[0108] n i is the propeller speed, where the power P L and torque T L of the AUV thruster are related as:

[0109] P Li = 2πn i T Li (1)

[0110] Determine the relationship between the propeller rotational speed and the thrust it generates, and establish the desired thrust and torque distribution formula:

[0111]

[0112] In the formula, D i is the propeller diameter of each thruster, ρ is the seawater density, K T and KQ is the torque correlation coefficient;

[0113] Combining formulas (1) and (2) gives:

[0114]

[0115] In the AUV multi-motor coordinated propulsion system, the relationship between the thrust and torque of the six-degree-of-freedom motion output of the AUV and the given thrust of the thrusters is established, and this relationship is used as the equality constraint condition in the AUV thrust allocation. The specific process includes:

[0116] A1: First, start the calculation of the combined thrust and torque of the horizontal thrusters. Define the thrust generated by each thruster as F i , according to the actual power layout of the No. 1 to No. 4 horizontal thrusters, where the No. 1 and No. 2 thrusters are symmetrically distributed about the X-axis to generate and balance the rotational torque; d xi and d yi respectively represent the longitudinal and lateral positions of each thruster on the AUV, and calculate the combined thrust and combined torque of the six-degree-of-freedom motion generated by the 4 horizontally arranged thrusters, which are expressed as:

[0117]

[0118] In the above formula represents the thrust vector of the single-target output of the 7 thrusters arranged on the AUV, is the resultant force and resultant torque received by the AUV in the horizontal direction;

[0119] A2: Perform the calculation of the combined thrust and torque of the vertical thrusters. For the convenience of calculation, consider the case where the vertically arranged thrusters are coplanar, and do not consider the interference of the horizontal thrusters on them. Here, the AUV is arranged with vertical thrusters in a triangular shape along the z-axis to meet the power supply; calculate the combined thrust and torque of the six-degree-of-freedom motion generated by the 3 vertically arranged thrusters, which are expressed as:

[0120]

[0121] In the above formula is the resultant force and resultant torque received by the AUV in the vertical direction;

[0122] A3: Finally, calculate the combined thrust and resultant torque of the 7 thrusters at different positions on the AUV propulsion system, and use the vector sum to obtain the combined thrust and resultant torque generated by the 4 horizontal thrusters and the 3 vertical thrusters, which are expressed as:

[0123] In the above formula [F X F Y F ZT K T M T N are the resultant force and resultant moment acting on the AUV;

[0124] Finally, the thrust and moment of the six-degree-of-freedom motion output of the AUV and the given thrust F of the thruster i The relationship is:

[0125]

[0126] Equation (6) is the equality constraint condition in the AUV thrust allocation. τ is the vector of the AUV output thrust and thrust moment, and B(α, β) is the AUV thruster vector arrangement matrix.

[0127] In step S3, based on the equality constraint condition in the AUV thrust allocation, the thrust control allocation problem of the AUV thruster is established, specifically:

[0128] B1: The AUV thrust allocation aims to minimize the energy consumption of the thrusters. According to the relationship between the power and thrust of the AUV thrusters, Equation (3) is further simplified to obtain:

[0129] P L = χF Ti 1.5 (7)

[0130] Where is a fixed constant, treated equivalently to ; For the convenience of subsequent calculations, the relationship between the power and thrust of the AUV thrusters is replaced by a quadratic form as Here, taking the minimum power consumption of the AUV thrusters as the optimization goal, the AUV thrust allocation objective function is established as:

[0131] min f = F Ti T MF Ti (8)

[0132] Where f is the power consumption of the AUV thrusters; F T is the thrust vector of the AUV thrusters, and M is a positive definite weighted coefficient matrix used to determine the weight of the AUV thruster energy consumption in the objective function;

[0133] B2: Continue to optimize the AUV thrust allocation error. Here, when the desired forces and moments of the AUV thrusters change greatly, the thrusters may not be able to adjust the rotational speed and azimuth angle in time to match the current required τ. To solve this problem, a slack variable s is introduced to ensure that the AUV thrust allocation problem has a feasible solution, expressed as:

[0134]

[0135] Among them, s is the difference between the control instruction generated by the AUV controller and the actual thrust after thrust allocation. If the difference is too large, it is impossible to ensure that the AUV sails on the expected trajectory or stays at the desired position. An AUV thrust allocation error function term is introduced:

[0136] J s = s T Qs (10)

[0137] Among them, Q is a positive definite weighting matrix used to ensure that the AUV thrust allocation error is controlled to zero;

[0138] B3: The AUV thruster thrust allocation equality constraint is expressed as The inequality constraint conditions consider the maximum thrust limit of the single-objective output of each thruster, the thrust change rate of the AUV thruster in a short time, and the thruster rudder angle change range limit, and are expressed as:

[0139]

[0140] Among them, F Ti is the output thrust magnitude of the i-th thruster on the AUV, F Timax and F Timin are the maximum and minimum values of the thrust output by the i-th thruster respectively; ΔF Ti is the thrust change rate of the AUV thruster in a short time, ΔF Timax and ΔF Timin are the maximum and minimum values of the change amount of the thrust of the i-th thruster in a short time respectively; α i is the rudder angle magnitude of the i-th thruster on the AUV, α imax and α imin are the maximum and minimum values of the rudder angle output by the i-th thruster respectively;

[0141] Finally, the objective function for the AUV thrust allocation problem is obtained as:

[0142]

[0143] The above inequality constraints include the thruster thrust change range constraint and the thruster saturation constraint, and the equality constraint realizes the controllability of the AUV attitude motion and the thrust error constraint.

[0144] The solution methods for solving the thrust control allocation problem by using the variable-objective AUV thrust allocation solution algorithm in step S4 include:

[0145] Case 1: When the AUV thruster thrust allocation is unsaturated or unconstrained, with the goal of minimizing energy consumption, the method of the solution space of the equations is used to obtain the optimal solution of the thrust allocation problem;

[0146] Case 2: When thrust saturation occurs, the method in Case 1 is changed to minimize the error, and the optimal solution of the thrust allocation problem is solved by the active set quadratic programming method;

[0147] Case 3: When a thruster fails, the optimal solution of the thrust allocation problem is directly solved by the active set quadratic programming method with the goal of minimizing the error of the AUV thruster thrust.

[0148] In Case 1, the specific solution process of the thrust control allocation problem is as follows:

[0149] C1: Adopt the thrust optimization allocation strategy based on the minimum energy consumption of the AUV thruster. According to the AUV thrust space layout matrix B(α,β), list the homogeneous equation system B i (α,β)F Ti = τ i , this equation system consists of FT0 and a particular solution FT*. The general solution is that the kernel of the AUV thrust space layout matrix B(α,β) is F T0 = Ker(B(α,β)), and the particular solution is directly obtained by the pseudo-inverse method as F T * = B(α,β) + τ, and the solution of the equation system is expressed as the sum of the general solution and the particular solution:

[0150] F Ti = χ i Ker(B i (α,β)) + B i (α,β) + τ (13)

[0151] where χ i is the set of valid solutions in the AUV thrust allocation problem, and the minimum energy consumption solution is found accordingly; write the inequality constraints of the general solution of the equation system according to formula (11):

[0152] F Timin - B i (α,β) + τ ≤ χKer(B i (α,β)) ≤ F Timax + B i (α,β) + τ (14)

[0153] When the χ i value exists, it indicates that there is no saturation problem in the thrust allocation, and the valid solution in the equation system is obtained by the pseudo-inverse method. Otherwise, if it does not exist, the quadratic programming method is used to solve the equation system;

[0154] C2: According to the quadratic model of formula (8), with the minimum energy consumption of AUV thrust allocation as the control optimization goal, the AUV thrust allocation objective function is expressed as:

[0155]

[0156] Taking the inner product of the general solution and the particular solution in formula (15), we get:

[0157] Ker(B i (α,β)) T B i (α,β) + τ=(B i (α,β)Ker(B i (α,β))) T (B(α,β)B(α,β) -1 ) T τ(16)

[0158] Since the kernel of the general solution is the zero vector, we get B i (α,β)Ker(B i (α,β))=0, and further calculation gives:

[0159] Ker(B i (α,β)) T B i (α,β) + τ=0 (17)

[0160] Under this condition, the minimum energy consumption problem is equivalently described as finding a solution that, while satisfying the given conditions, minimizes the energy consumption. Its optimal energy consumption problem is expressed as:

[0161]

[0162] Equivalent the above minimum energy consumption problem to the value problem of χ i When χ i ≠0, the minimum energy consumption solution of AUV thrust allocation is expressed as:

[0163]

[0164] When χ i =0, the minimum energy consumption solution of AUV thrust allocation is solved by the pseudoinverse method and expressed as:

[0165]

[0166] The above is the particular solution of the system of equations space method and can only be solved when the AUV thruster thrust allocation is not saturated or unconstrained.

[0167] In Case 2, the specific solution process of the thrust control allocation problem is as follows:

[0168] The optimization allocation based on minimizing the thrust error of the AUV thruster is adopted. Since the space method of the above equations cannot solve the case of AUV thruster thrust saturation, the active set in quadratic programming is used to find the optimal solution when the thrust is saturated. The AUV thrust optimization allocation problem is expressed in quadratic programming as:

[0169]

[0170] In the formula, W in the control objective function is an n-order symmetric matrix, which is determined by the quadratic coefficient in the objective function; E and G are subsets of equality constraints and inequality constraints respectively, and A, B, C, and s are also n-order vectors;

[0171] Then, it is rewritten into an incremental form through the Taylor formula as:

[0172]

[0173] The above optimal effective particular solution of AUV thrust allocation needs to satisfy the KT (Kuhn Tucker) condition:

[0174]

[0175] Among them, is the search gradient of the Lagrangian function in the AUV thrust allocation optimization problem, and h(s * )≥0 represents the inequality constraint imposed on the AUV thrust optimization allocation, and it is a necessary and sufficient condition for s * to be the effective solution of the equation system; when χ i ≥0 is the dual satisfaction, we get When χ i <0, it means that the search gradient descent direction of the control objective function is consistent with the gradient direction of the constraint boundary. Therefore, the obtained solution is not the optimal solution, that is, it does not satisfy the dual effectiveness, and s * is not the optimal solution of the thrust allocation problem; when the optimal solution is inside the control domain, that is, h(s * )≥0, it is equivalent to an unconstrained optimization problem. At this time, χ i =0; on the contrary, if the optimal solution in the thrust allocation problem is located on the boundary of the control domain, that is, h(s * )=0, then the optimal solution problem must satisfy that the product of χi and h(s * ) is zero, that is, the complementary slackness;

[0176] The AUV thrust optimization problem is transformed into an optimization solution using the active set algorithm in quadratic programming. The active set method includes initializing the need to provide an initial feasible solution s0 and an initial working set G0, taking the optimal solution in the thrust allocation problem as the initial solution of the optimization problem, and setting the initial set as an empty set; first, find the search direction δ k = s k - s k-1 and the corresponding Lagrange multiplier η i . The iteration step size is defaulted to O k = 1; then transform the inequality constraints in the working set G0 into equality constraints that only restrict the boundary conditions, solve for the global minimum solution s’ in its sub-problem, and set the corresponding Lagrange multiplier η i ; if |s' - s k | < ε, where ε is an infinitely small number greater than zero, and χ i ≥ 0, then the optimal solution of the AUV thrust allocation problem is obtained. Update the corresponding active set according to the sign of the Lagrange multiplier χ i , and remove the ineffective constraint opportunities from G k * ; if the new feasible solution is not in the feasible domain, select an appropriate step size through the current search direction to bring it back to the feasible domain. The process is expressed by the formula:

[0177]

[0178] If |s' - s k | > ε, then s’ is a solution on the feasible domain of the thrust allocation problem, and then s k+1 = s'. Through the above formula iteration update, if s’ is not a solution on the feasible domain, according to formula (1.25), the following assumptions are made:

[0179]

[0180] Its iteration rule is:

[0181]

[0182] Finally, based on the AUV thruster power layout and mutual constraint parameters, combined with the minimization of the AUV thruster thrust allocation optimization method, the optimization model is obtained as:

[0183]

[0184] When the thruster fails in step S4, that is, in case three, the originally possible fully actuated thrust configuration may change to an underactuated state. In this case, the thrust allocation problem may become unsolvable. To address this situation, the effective set method can be directly used to solve the optimal solution with the goal of minimizing the error of the AUV thruster thrust:

[0185] Remove the inequality constraints of the AUV thruster in the following formula:

[0186] F Timin ≤F Ti ≤F Timax (28)

[0187] And add the equality constraints for the corresponding failed thruster:

[0188] F Ti =0 (29)

[0189] Finally, construct an inequality quadratic programming problem to solve the fault tolerance problem of the AUV thrust allocation problem. The specific working process of its variable target thrust optimization allocation is as Figure 3 shown. It can be described as follows: First, judge whether the AUV thruster fails according to the previous thrust control command of the AUV thruster. If a failure occurs, directly use the AUV thruster thrust optimization allocation method with minimized error. Otherwise, use the AUV thruster thrust optimization allocation method based on minimum energy consumption, and further judge whether the AUV thruster thrust output is saturated. If there is no saturation, directly output the AUV thruster thrust. Otherwise, perform the AUV thruster thrust error optimization allocation based on error minimization again.

[0190] In the present invention, a variable target thrust optimization allocation method is proposed to achieve the scenario adaptive balance between energy consumption and accuracy, thereby realizing the precise allocation of the AUV thruster thrust. By designing an accurate AUV thruster model and determining the AUV thrust allocation process to ensure the power source for AUV underwater operations, then determining the power layout of the AUV propulsion system and the thruster thrust control allocation problem, and finally proposing a variable target AUV thruster thrust allocation method. When there is no thrust saturation in the AUV thruster thrust allocation, the minimum energy consumption thrust allocation with good real-time performance is adopted; when thrust saturation occurs, switch to the minimum error allocation strategy. It provides a new solution for the application of the AUV propulsion system in the coordinated operation of multiple propulsion motors.

[0191] The present invention constructs an accurate thruster model and determines the AUV thrust allocation process, further builds the power layout of the AUV propulsion system to facilitate the AUV to complete six-degree-of-freedom spatial motion, analyzes the basic problems of AUV thruster thrust allocation based on the AUV power layout, and proposes a thrust allocation algorithm composed of variable objectives combining the minimum energy consumption and minimum error of the AUV. The present invention adopts a thrust allocation algorithm composed of variable objectives of minimum energy consumption and minimum error, which combines the advantages of the pseudo-inverse method and the quadratic programming method, and can flexibly adjust the optimization method under different working conditions. This method aims to dynamically adjust the optimization objective according to the real-time navigation requirements to achieve the optimal compromise between energy consumption and accuracy.

Claims

1. An underwater thruster thrust distribution method based on multi-motor cooperation, characterized in that The steps are as follows: S1: The desired attitude given by the AUV multi-motor coordinated propulsion system and the actual feedback attitude of the AUV are input into the AUV attitude controller to obtain the control force signal; S2: The thrust distributor decomposes the control force signal in real time and converts it into a rotational speed signal to be transmitted to the AUV multi-motor coordinated propulsion system; S3: Establish the thrust control allocation problem of the AUV thruster; S4: Use the variable-target AUV thrust allocation solution algorithm to solve the thrust control allocation problem and output the thruster thrust allocation result.

2. A thrust distribution method for an underwater thruster based on multi-motor cooperation according to claim 1, characterized in that In step S1, the AUV multi-motor coordinated propulsion system is equipped with multiple thrusters, and their layout and design are as follows: The number of thrusters is 7, namely thruster No. 1 to No.

7. Thruster No. 1 and No. 2 are propeller-rudder combined thrusters of the AUV. They are arranged along the x-axis direction. The dynamic control of the thrust direction of a single AUV thruster is achieved through rudder angle adjustment. Their control angles α1, α2 ∈ [-90°, 90°] and the clockwise direction is positive; Thruster No. 3 and No. 4 are fixed-angle side thrusters, symmetrically distributed with the x-axis and the included angle is taken as β1 = β2 = 45°, providing lateral thrust compensation for the AUV; Thruster No. 1 to No. 4 are all horizontal thrusters, providing X-Y-N degree-of-freedom control forces and torques for the AUV propulsion system; Thruster No. 5 to No. 7 are vertical thrusters and are arranged in a triangle along the z-axis. The combined action of these three vertical thrusters provides Z-M-K degree-of-freedom control forces and torques for the AUV.

3. A thrust distribution method for an underwater thruster based on multi-motor cooperation according to claim 2, characterized in that In the AUV multi-motor coordinated propulsion system, a mathematical model is established for a single thruster: n i is the propeller speed, where the AUV thruster power P L and torque T L are related as follows: P Li = 2πn i T Li (1) Determine the relationship between the propeller rotational speed and the thrust generated by it, and establish the desired thrust and torque allocation formula: where D i is the diameter of each thruster propeller, ρ is the seawater density, K T and K Q are moment-related coefficients; Combining formulas (1) and (2) gives:

4. A thrust allocation method for an underwater thruster based on multi-motor cooperation according to claim 2, wherein In the AUV multi-motor coordinated propulsion system, establish the relationship between the thrust and torque of the AUV six-degree-of-freedom motion output and the given thrust of the thruster, and use this relationship as the equality constraint condition in the AUV thrust allocation. The specific process includes: A1: First, start with the calculation of the combined thrust and moment of the horizontal thrusters. Define the thrust generated by each thruster as F i , according to the actual power layout of thrusters 1 to 4 in the horizontal direction. Among them, thrusters 1 and 2 are symmetrically distributed about the X-axis to generate and balance the rotational moment; d xi and d yi respectively represent the longitudinal and lateral positions of each thruster on the AUV. Calculate the magnitudes of the combined thrust and combined moment of the six-degree-of-freedom motion generated by the four horizontally arranged thrusters, which are expressed as: In the above formula represents the thrust vectors of the single-objective outputs of the seven thrusters arranged on the AUV, which are the resultant force and resultant moment acting on the AUV in the horizontal direction; A2: Conduct the calculation of the combined thrust and torque of the vertical thrusters. The vertical thrusters of the AUV arranged in a triangle along the z-axis are used to meet the power given; Calculate the magnitudes of the combined thrust and torque of the six-degree-of-freedom motion generated by the 3 vertically arranged thrusters, expressed as: In the above formula is the resultant force and resultant moment acting on the AUV in the vertical direction; A3: Finally, calculate the combined thrust and combined torque of the 7 thrusters at different positions on the AUV propulsion system, and use the vector sum to obtain the combined thrust and combined torque generated by the 4 horizontal thrusters and the 3 vertical thrusters, expressed as: In the above formula, [F X F Y F Z T K T M T N are the resultant force and resultant moment received by the AUV; Finally, the thrust and moment of the six-degree-of-freedom motion output of the AUV and the given thrust F of the thruster i The relationship is: Formula (6) is the equality constraint condition in the AUV thrust allocation. τ is the AUV output thrust and thrust moment vector, and B(α,β) is the AUV thruster vector layout matrix.

5. A thrust allocation method for an underwater thruster based on multi-motor coordination according to claim 4, characterized in that, In step S3, based on the equality constraint condition in the AUV thrust allocation, establish the thrust control allocation problem of the AUV thruster, specifically: B1: The AUV thrust allocation aims to minimize the energy consumption of the thrusters. According to the relationship between the power and thrust of the AUV thrusters, formula (3) is further simplified to obtain: P L = χF Ti 1.5 (7) Among them, is a fixed constant and is treated equivalently to Regarding the relationship between the power and thrust of the AUV thruster, it is replaced by a quadratic form as Taking the minimum power consumption of the AUV thruster as the optimization goal, the AUV thrust distribution objective function is established as: minf = F Ti T MF Ti (8) where f is the power consumption of the AUV thruster; F T is the thrust vector of the AUV thruster, and M is a positive definite weighted coefficient matrix used to determine the weight of the AUV thruster energy consumption in the objective function; B2: Continue to optimize the AUV thrust allocation error. Introduce the slack variable s to ensure that the AUV thrust allocation problem has a feasible solution, expressed as: Among them, s is the difference between the control instruction generated by the AUV controller and the actual thrust after thrust allocation. If the difference is too large, it is impossible to ensure that the AUV sails on the expected trajectory or stays at the desired position. An AUV thrust allocation error function term is introduced: J s = s T Qs (10) Among them, Q is a positive definite weighting matrix used to ensure that the AUV thrust allocation error is controlled to zero; B3: For the equality constraint of AUV thruster thrust distribution The inequality constraint conditions consider the maximum thrust limit of the single-objective output of each thruster, the thrust change rate of the AUV thruster in a short time, and the restriction of the thruster rudder angle change range, and are expressed as: Among them, F Ti is the magnitude of the thrust output by the i-th thruster on the AUV, F Timax and F Timin are respectively the maximum and minimum values of the thrust output by the i-th thruster; ΔF Ti is the change rate of the thrust of the AUV thruster in a short time, ΔF Timax and ΔF Timin are respectively the maximum and minimum values of the change amount of the thrust of the i-th thruster in a short time; α i is the magnitude of the rudder angle of the i-th thruster on the AUV, α imax and α imin are respectively the maximum and minimum values of the rudder angle output by the i-th thruster; Finally, the objective function of the AUV thrust allocation problem is obtained as: minF = F Ti T MF Ti +s T Qs The inequality constraints in the above formula include the constraints on the range of thruster thrust variation and the thruster saturation constraints, and the equality constraints realize the controllability of the AUV attitude motion and the thrust error constraints.

6. A thrust distribution method for an underwater thruster based on multi-motor cooperation according to claim 5, characterized in that, The solution method for solving the thrust control allocation problem by using the variable-objective AUV thrust allocation solution algorithm in step S4 includes: Case 1: When the AUV thruster thrust allocation is not saturated or unconstrained, with the goal of minimizing energy consumption, the method of the solution space of the equations is used to obtain the optimal solution of the thrust allocation problem; Case 2: When thrust saturation occurs, the method changes from that in Case 1 to the goal of minimizing the error, and the optimal solution of the thrust allocation problem is solved by the active set quadratic programming method; Case 3: When a thruster fails, directly with the goal of minimizing the error of the AUV thruster thrust, the optimal solution of the thrust allocation problem is solved by the active set quadratic programming method.

7. A thrust distribution method for an underwater thruster based on multi-motor cooperation according to claim 6, characterized in that In Case 1 of step S4, the specific solution process of the thrust control allocation problem is: C1: Adopt a thrust optimization distribution strategy for AUV based on minimum energy consumption. According to the AUV thrust space layout matrix B(α,β), list the homogeneous equations related to AUV thrust distribution B i (α,β)F Ti =τ i , this system of equations consists of FT0 and a particular solution FT*. The general solution is that the kernel of the AUV thrust space layout matrix B(α,β) is F T0 =Ker(B(α,β)), and the particular solution is directly obtained by the pseudoinverse method as F T * =B(α,β) + τ, and the solution of the system of equations is expressed as the sum of the general solution and the particular solution: F Ti = χ i Ker(B i (α,β)) + B i (α,β) + τ (13) where χ i is the set of valid solutions for the AUV thrust allocation problem, and the minimum energy consumption solution is found accordingly; write the inequality constraints for the general solution of the system of equations according to formula (11): F Timin -B i (α,β) + τ ≤ χKer(B i (α,β)) ≤ F Timax +B i (α,β) + τ (14) When χ i exists, it indicates that there is no saturation problem in thrust allocation, and the effective solution in the equation set is obtained by the pseudo-inverse method. Otherwise, the quadratic programming method is used to solve the equation set; C2: According to the quadratic model of formula (8), with the minimum energy consumption of AUV thrust allocation as the control optimization goal, the AUV thrust allocation objective function is expressed as: Taking the inner product of the general solution and the particular solution of the equation in formula (15), we get: Ker(B i (α,β)) T B i (α,β) + τ=(B i (α,β)Ker(B i (α,β))) T (B(α,β)B(α,β) -1 ) T τ(16) Since the kernel of the general solution is the zero vector, we obtain B i (α,β)Ker(B i (α,β)) = 0, and further calculations yield: Ker(B i (α,β)) T B i (α,β) + τ=0 (17) Under this condition, the minimum energy consumption problem is equivalently described as finding a solution, which, while satisfying the given conditions, minimizes the energy consumption. Its optimal energy consumption problem is expressed as: The above minimum energy consumption problem is equivalent to the problem of the value of χ i When χ i ≠ 0, the minimum energy consumption solution of AUV thrust allocation is expressed as: When χ i = 0, the minimum energy consumption solution of AUV thrust allocation is solved by the pseudo-inverse method and expressed as: The above formula is the particular solution of the equations space method and can only be solved when the AUV thruster thrust allocation is not saturated or unconstrained.

8. A thrust allocation method for an underwater thruster based on multi-motor cooperation according to claim 6, characterized in that, In Case 2 of step S4, the specific solution process of the thrust control allocation problem is: When thrust saturation occurs, the active set in the quadratic programming method is used to find the optimal solution. The AUV thrust optimization allocation problem is expressed in quadratic programming as: In the formula, W in the control objective function is an n-order symmetric matrix determined by the quadratic coefficient in the objective function; E and G are the subsets of equality constraints and inequality constraints respectively, and A, B, C, and s are also n-order vectors; Then, it is rewritten into an incremental form through the Taylor formula as: The above optimal active particular solution of AUV thrust allocation needs to satisfy the KT condition: Among them, is the search gradient of the Lagrangian function in the AUV thrust allocation optimization problem, and h(s * ) ≥ 0 represents the inequality constraint imposed on the AUV thrust optimization allocation, and it is a necessary and sufficient condition for s * to be an effective solution of the system of equations; when χ i ≥ 0 is the dual feasibility, we get while when χ i < 0, it means that the search gradient descent direction of the control objective function is consistent with the gradient direction of the constraint boundary. Therefore, the obtained solution is not the optimal solution, that is, it does not satisfy the dual effectiveness, and s * is not the optimal solution to the thrust allocation problem; when the optimal solution is inside the control domain, that is, h(s * ) ≥ 0, it is equivalent to an unconstrained optimization problem. At this time, χ i = 0; on the contrary, if the optimal solution in the thrust allocation problem is located on the boundary of the control domain, that is, h(s * ) = 0, then this optimal solution problem must satisfy that the product of χ i and h(s * ) is zero, that is, the complementary slackness; The AUV thrust optimization problem is transformed into an active set algorithm in quadratic programming for optimization and solution. The active set method includes initializing the need to provide an initial feasible solution s0 and an initial working set G0, taking the optimal solution in the thrust allocation problem as the initial solution of the optimization problem, and setting the initial set as an empty set; first, find the search direction δ k = s k - s k-1 and the corresponding Lagrange multiplier η i , and the iteration step size is defaulted to O k = 1; then transform the inequality constraints in the working set G0 into equality constraints that only restrict the boundary conditions, solve the global minimum solution s' in its sub-problem, and set the corresponding Lagrange multiplier η i ; if |s' - s k | < ε, where ε is an infinitely small number greater than zero, and χ i ≥ 0, then the optimal solution of the AUV thrust allocation problem is obtained. Update the corresponding active set according to the sign of the Lagrange multiplier χ i , and remove the ineffective constraint opportunities from G k * ; if the new feasible solution is not in the feasible region, select an appropriate step size through the current search direction to make it return to the feasible region. The process is expressed by the formula: If |s' - s k | > ε, then s' is a solution on the feasible region in the thrust allocation problem, and s k+1 = s'. Through the iterative update of the above formula, if s' is not a solution on the feasible region, the following assumptions are made according to formula (1.25): Its iteration rule is: Finally, based on the AUV thruster power layout and the mutual constraint parameters, combined with the method of minimizing the AUV thruster thrust allocation optimization, the optimization model is obtained as:

9. A thrust distribution method for an underwater thruster based on multi-motor cooperation according to claim 8, characterized in that, When a thruster fails in step S4, directly with the goal of minimizing the error of the AUV thruster thrust, the active set method is used to solve the optimal solution: Remove the inequality constraints of the following AUV thruster: F Timin ≤ F Ti ≤ F Timax (28) And add the equality constraints of the corresponding failed thruster: F Ti ＝0 (29)。

Images