Underwater vehicle linear hydrodynamic parameter identification method for control law design

Abstract

The application discloses a kind of linear hydrodynamic parameter identification methods of underwater vehicle facing control law design, it is related to the cross field of ship integrated motion control technology and bionic intelligent algorithm, including the following steps: establishing linear motion state equation of underwater vehicle;Obtain underwater vehicle navigation test data;Determine objective function in combination with underwater vehicle hydrodynamic motion characteristics;The hydrodynamic parameters contained in linear motion state equation are taken as individuals to form a population, and the population is initialized;According to the objective function, the population is optimized using improved genetic algorithm, and the key hydrodynamic parameters are output.The application effectively reduces the error between the output of the linear motion equation of the underwater vehicle under typical steering input and the test data, thereby reducing the modeling error of the linear motion equation of the underwater vehicle used for controller design, and improving the control quality and robustness of the control algorithm of the underwater vehicle.

Description

Technical Field

[0001] The present invention relates to the field of ship integrated motion control technology and bionic intelligent algorithm technology, and more specifically to a method for identifying linear hydrodynamic parameters of underwater vehicles oriented to control law design. Background Art

[0002] The mission of an underwater vehicle's maneuvering control system is to ensure that the vehicle meets the required heading, depth, and attitude during normal navigation. The control laws involved in this system are crucial to the successful execution of this mission and the safety of navigation operations. Advanced underwater vehicle technology countries like the United States, Russia, and Germany all use model-based control theory to develop control laws for underwater vehicle maneuvering control systems.

[0003] Currently, my country's underwater vehicles primarily use traditional PID control methods to design control laws. While this method effectively controls heading and depth, it lacks precise and effective control of heading rate and attitude, such as pitch. Model-based control law design can effectively control the attitude of underwater vehicles. By identifying and correcting the key hydrodynamic coefficients involved in the control law design process, the control quality of the underwater vehicle's pitch and heading rate can be further improved.

[0004] Therefore, how to determine the key parameters of underwater vehicles and improve the control quality of the pitch and heading rate control of underwater vehicles is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides an underwater vehicle linear hydrodynamic parameter identification method for control law design, which ensures the robustness and control quality of the underwater vehicle control algorithm and meets the requirements of underwater vehicle navigation attitude rate control law design.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for identifying the linear hydrodynamic parameters of an underwater vehicle for control law design includes the following steps:

[0008] Step 1: Establish the linear motion state equation of the underwater vehicle:

[0009] Step 2: Acquire the test data of the underwater vehicle within the next time period at the preset rudder angle;

[0010] Step 3: constructing an objective function based on the linear motion state equation of the underwater vehicle and the test data; the objective function is the deviation between the deviation of the linear motion state equation process simulation and the actual heading, pitch, and depth in the test data;

[0011] Step 4, the water dynamic parameters contained in the linear motion state equation in step 1 are taken as individuals to form a population, and the population is initialized;

[0012] Step 5, the population is optimized by using the improved genetic algorithm according to the target function, and the key water dynamic parameters are output.

[0013] Preferably, the test data include longitudinal velocity u, depth ζ, heading pitch θ, rudder angle δ r , bow rudder angle δ b , stern rudder angle δ s , and the like.

[0014] Preferably, the linear motion state equation in step 1 includes a horizontal plane linear motion state equation and a vertical plane linear motion state equation, which are respectively represented as:

[0015] Horizontal plane linear motion state equation:

[0016]

[0017]

[0018] Vertical plane linear motion state equation:

[0019]

[0020]

[0021] Wherein, v, w, p, q, r are respectively the lateral velocity, vertical velocity, roll angular velocity, pitch angular velocity and yaw angular velocity in the ship body coordinate system; θ, ζ are respectively the pitch angle, heading angle and vertical displacement in the fixed coordinate system; δ r , δ b , δ s are respectively the rudder angle, bow rudder angle and stern rudder angle; A 11 , A 12 , A 21 , A 22 , B1, B2 are the key water dynamic parameters affecting the steering control in the horizontal plane linear motion state equation; a 11 , a 12 , a 13 , a 21 , a 22 , a 23 , b 11 , b 12 , b 21 , b 22 are the key water dynamic parameters affecting the steering control in the vertical plane linear motion state equation.

[0022] Preferably, the objective function constructed in step 3 includes a horizontal plane motion identification objective function S1 and a vertical plane motion identification objective function S2, respectively represented as:

[0023]

[0024]

[0025] wherein n is the number of data in the time period t, is the test heading data, is the maximum heading value in the time period t, is the minimum heading value in the time period t, is the test pitch data, θ max is the maximum pitch value in the time period t, θ min is the minimum pitch value in the time period t, is the test depth data, ζ max is the maximum depth value in the time period t, ζ min is the minimum depth value in the time period t.

[0026] Preferably, the initialization process of the population in step 4 includes:

[0027] setting the total number of the population as M, and randomly generating each individual in the population within a preset range according to the theoretical value of the hydrodynamic coefficient of the underwater vehicle corresponding to the shape size experience; the preset range is the theoretical value of the hydrodynamic coefficient ± 30%.

[0028] Preferably, the step of optimizing the population by using the improved genetic algorithm in step 5 includes:

[0029] Step 51, performing evolution operation on the population;

[0030] Step 52, performing mutation operation on the population:

[0031] Step 53, defining the maximum number of iterations, repeating step 51 and step 52 until the maximum number of iterations is met, and outputting the corresponding global optimal individual, the global optimal individual as the key hydrodynamic parameter.

[0032] Preferably, the evolution operation process of the population in step 51 is as follows:

[0033] Step 511: calculating the objective function value corresponding to each individual in the population, forming an objective function value vector J, represented as:

[0034] J = [J1 … J i … J M]' (7)

[0035] wherein J i is the objective function value of the ith individual, and M is the size of the population;

[0036] Step 512: sort the objective function values in the objective function value vector in descending order, and mark the corresponding individual with the sorting position, and set the selection pressure sp;

[0037] Step 513: calculate the fitness of the individual according to the sorting position and the selection pressure, and the expression is as follows:

[0038]

[0039] fiti represents the fitness of the ith individual, pos i represents the corresponding sorting position of the ith individual, and sp represents the selection pressure; then the corresponding selection probability Pi is as follows:

[0040]

[0041] Step 514: perform a crossover operation on the individual, and the calculation method is as follows:

[0042]

[0043] wherein, and respectively represent two individuals performing the crossover operation, respectively represent the corresponding new individuals after the crossover operation of the two individuals; a is a vector parameter with the same dimension as the individual, and each component of a is a uniform random number in the interval [-d, 1+d], and d is a constant.

[0044] Preferably, in step 5, the mutation operation is performed on the individual to obtain the optimal solution. Random mutation of the individual's dye improves diversity, avoids premature local convergence, and finds the global optimal solution; Gaussian mutation of the population means that one or more original gene values at the chromosome locus are replaced by a random number conforming to a normal distribution with a mean of and a variance of σ 2 Gaussian mutation can focus on searching the local area to find the optimal solution; if the value to be mutated is X, a Gaussian random number with a mean of X and a variance of 0.05 times the length of the boundary range can be generated to replace X; applying Gaussian mutation can also cause the mutated value to exceed the boundary, so as to ensure that the offspring obtained after mutation is within the boundary range.

[0045] Compared with the prior art, the underwater vehicle linear hydrodynamic parameter identification method for control law design provided by the present application adopts an improved genetic algorithm to identify key hydrodynamic parameters affecting the control law design of the underwater vehicle, and a target function is constructed by fitting the simulation curve of the linear state motion equation of the underwater vehicle with the underwater vehicle navigation curve, so as to improve the similarity of the state equation and the underwater vehicle. The method only needs to identify a small amount of hydrodynamic parameters, can avoid the identification result from falling into a local optimal solution, improves the identification accuracy, and provides a way to eliminate the influence of modeling errors caused by the scale effect on the control law design, so as to further improve the comprehensive control quality of the underwater vehicle control system and the robustness of the control algorithm. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the provided drawings.

[0047] Figure 1 The underwater vehicle linear hydrodynamic parameter identification method for control law design provided by the present application is shown in the flowchart.

[0048] Figure 2 The simulation curve of the state equation before identification provided by the present application is shown in the schematic diagram.

[0049] Figure 3 The simulation curve of the state equation after identification provided by the present application is shown in the schematic diagram.

[0050] Figure 4 The comparison schematic diagram of the control test data of the underwater vehicle before identification provided by the present application is shown in the schematic diagram.

[0051] Figure 5 The comparison schematic diagram of the control test data of the underwater vehicle after identification of the key hydrodynamic parameters provided by the present application is shown in the schematic diagram. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0053] The embodiment of the present application discloses an underwater vehicle linear hydrodynamic parameter identification method for control law design, which comprises the following steps: Figure 1As shown, it is a flow chart of an adaptive mutation particle swarm identification algorithm, comprising the following steps:

[0054] S1: The design of the control law divides the underwater vehicle motion into horizontal plane motion and vertical plane motion, and constructs the linear motion state equation of the underwater vehicle; wherein:

[0055] The horizontal plane linear motion state equation is:

[0056]

[0057]

[0058] The vertical plane linear motion state equation is:

[0059]

[0060]

[0061] Wherein, v, w, p, q, r are the lateral velocity, vertical velocity, roll angle velocity, pitch angle velocity, yaw angle velocity in the ship body coordinate system; θ, ζ are the pitch angle, heading angle, vertical displacement in the fixed coordinate system; δ r , δ b , δ s are the rudder angle, bow rudder angle, stern rudder angle; A 11 , A 12 , A 21 , A 22 , A 32 , B1, B2 are the key hydrodynamic parameters affecting the steering control in the horizontal plane state space equation, a 11 , a 12 , a 13 , a 21 , a 22 , a 23 , a 32 , a 41 , a 43 , b 11 , b 12 , b 21 , b 22 are the key hydrodynamic parameters affecting the steering control in the vertical plane state space equation; from the linear state space equation of the underwater vehicle, when the longitudinal velocity u, depth ζ, heading pitch θ, rudder angle δ r , bow rudder angle δ b , stern rudder angle δ s and other data of the underwater vehicle are known, there must be a set of suitable hydrodynamic parameters to make the simulation results after identification and the test data have the highest fitting degree;

[0062] S2: Collecting the data of longitudinal velocity u, depth ζ, heading pitch θ, rudder angle δ r , bow rudder angle δ b , stern rudder angle δ s of underwater vehicle in a time t under a given rudder angle;

[0063] S3: Determining the objective function of underwater vehicle motion control according to the data collected in S2 and the data simulated in S1 linear motion state equation; wherein:

[0064] Horizontal plane motion identification objective function S1:

[0065]

[0066] Vertical plane motion identification objective function S2:

[0067]

[0068] Wherein n is the number of data in time t, is the test heading data, is the maximum heading value in time t, is the minimum heading value in time t, is the test pitch data, θ max is the maximum pitch value in time t, θ min is the minimum pitch value in time t, ζ i * is the test depth data, ζ max is the maximum depth value in time t, ζ min is the minimum depth value in time t; The reason for selecting the normalized mean square error as the objective function is that the vertical plane motion mainly considers the weight inconsistency of the two states of depth and pitch, and the normalized mean square error can balance the influence of depth and pitch error on the objective function;

[0069] S4: Initializing the population;

[0070] The total number of horizontal plane motion population individuals is 50, the individual chromosome dimension is 6, 50 6-dimensional arrays are generated, and each value in the array is randomly generated within ± 30% of the water dynamic coefficient theoretical value obtained from the corresponding shape size experience; The total number of vertical plane motion particle population individuals is 100, the individual chromosome dimension is 10, 100 10-dimensional arrays are generated, and each value in the array is within ± 30% of the water dynamic coefficient theoretical value obtained from the corresponding shape size experience.

[0071] S5: Evolution operation is performed on the particles;

[0072] S51: After calculating the objective function of each individual in the population, use vector J to represent the objective function value of each individual in the population;

[0073] J=[J1 … J i … J M ]' (7)

[0074] Among them J i is the objective function value of the i-th individual, M is the size of the population; if the i-th individual is sorted in descending order by the objective function value, the corresponding sorted position is pos i , the selection pressure is set to sp.

[0075] S52: The fitness of an individual is calculated as follows:

[0076]

[0077] The fitness of the i-th individual is fiti, so its corresponding probability of being selected Pi is as follows:

[0078]

[0079] S53: Perform cross manipulation on individuals. Assume that the two individuals to be cross manipulated are and The calculation method of the new dyed individual after the crossover operation is as follows:

[0080]

[0081] Where α is a vector parameter with the same dimension as the individual, and the value of each component is a uniform random number in the interval [-d, 1+d]; d is taken as 0.25;

[0082] S6: Perform mutation operation on the population:

[0083] Perform Gaussian mutation on the population. Gaussian mutation means using the mean value to mutate the population. The variance is σ 2 A random number from a normal distribution replaces one or more original gene values ​​on a chromosome locus with a small probability. Gaussian mutation can focus on searching local areas to find the optimal solution. If the value to be mutated is X, a Gaussian random number with a mean of X and a variance of 0.05 times the length of the boundary range can be generated to replace X. Applying Gaussian mutation may also cause the mutated value to exceed the boundary. To ensure that the offspring obtained after mutation are within the boundary range, the value exceeding the boundary is modified to the nearest boundary.

[0084] S7: Determine the maximum number of iterations to be 100, repeat S5 and S6 until the maximum number of iterations is met, and output the global best position.

[0085] Figure 2 and Figure 4 For the identification of the front underwater vehicle control state equation simulation curve and the test data comparison chart, Figure 3 and Figure 5 For the identification of the front underwater vehicle control state equation simulation curve and the test data comparison chart, The results show that the identification method of the present application can improve the similarity of the linear motion equation of the underwater vehicle and the motion of the underwater vehicle, improve the modeling accuracy of the linear motion equation of the underwater vehicle, and provide an efficient key hydrodynamic coefficient identification algorithm for the development of the control law of the underwater vehicle. It can be widely applied in the field of aviation for the development of the flight control system control law.

[0086] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0087] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for identifying linear hydrodynamic parameters of underwater vehicles for control law design, characterized by: The specific steps include: Step 1: Establish the linear motion state equation of the underwater vehicle: Step 2: Acquire the test data of the underwater vehicle within the next time period at the preset rudder angle; Step 3: constructing an objective function based on the linear motion state equation of the underwater vehicle and the test data; Step 4: Use the hydrodynamic parameters contained in the linear motion state equation in step 1 as individuals to form a population and initialize the population; Step 5: Optimize the population using an improved genetic algorithm based on the objective function and output key hydrodynamic parameters; The linear motion state equations in step 1 include the horizontal plane linear motion state equations and the vertical plane linear motion state equations, which are respectively expressed as: State equation of linear motion on the horizontal plane: State equation of linear motion in the vertical plane: Among them, v, w, p, q, and r are the lateral velocity, vertical velocity, heel angular velocity, pitch angular velocity, and yaw angular velocity in the ship's hull coordinate system respectively; θ, ζ are the pitch angle, heading angle and vertical displacement in the fixed coordinate system respectively; δ r , δ b , δ s are the rudder angle, bow rudder angle and stern rudder angle respectively; A 11 、A 12 、A 21 、A 22 , B1, B2 are the key hydrodynamic parameters in the horizontal plane linear motion state equation, a 11 、a 12 、a 13 、a 21 、a 22 、a 23 、b 11 、b 12 、b 21 、b 22 is the key hydrodynamic parameter in the vertical plane linear motion equation of state; The objective function constructed in step 3 includes a horizontal plane motion identification objective function S1 and a vertical plane motion identification objective function S2, which are respectively expressed as: Among them, n is the number of data in time period t, For the test heading data, is the maximum heading value in time period t, is the minimum heading value in time period t, θ i * is the test pitch data, θ max is the maximum pitch value in time period t, θ min is the minimum pitch value in time period t, ζ i * is the test depth data, ζ max is the maximum depth value in time period t, ζ min is the minimum depth value in time period t.

2. The method for identifying linear hydrodynamic parameters of underwater vehicles for control law design according to claim 1 is characterized in that: The test data include longitudinal velocity u, depth ζ, heading Pitch θ, rudder angle δ r , bow rudder angle δ b and stern rudder angle δ s .

3. The method for identifying linear hydrodynamic parameters of underwater vehicles for control law design according to claim 1 is characterized in that: When initializing the population in step 4, the total population is set to M, and a setting range is preset according to the theoretical value of the hydrodynamic coefficient corresponding to the underwater vehicle, and each individual in the population is randomly generated within the setting range.

4. The method for identifying linear hydrodynamic parameters of underwater vehicles for control law design according to claim 1 is characterized in that: The step of optimizing the population using the improved genetic algorithm in step 5 includes: Step 51: performing an evolution operation on the population; Step 52: Perform mutation operation on the population: Step 53: define a maximum number of iterations, repeat steps 51 and 52 until the maximum number of iterations is met, and output the global optimal individual, which serves as the key hydrodynamic parameter.

5. The method for identifying linear hydrodynamic parameters of underwater vehicles for control law design according to claim 1 is characterized in that: The process of performing the evolution operation on the population in step 51 includes: Step 511: Calculate the objective function value corresponding to each individual in the population to form an objective function value vector J, which is expressed as: J=[J1…J i …I M ]' Among them, J i is the objective function value of the i-th individual, and M is the size of the population; Step 512: sort the objective function values ​​in descending order and set the selection pressure to sp; Step 513: Calculate the fitness of the individual based on the ranking position and selection pressure. The expression is as follows: fit i represents the fitness of the i-th individual, pos i represents the ranking position of the i-th individual after sorting, sp is the selection pressure; the probability P of the i-th individual being selected i As shown in the following formula: Step 514: Perform a crossover operation on the individuals. The calculation formula is: in, and Represent the two individuals performing the crossover operation, They represent the new individuals corresponding to the crossover operation of two individuals; α is a vector parameter with the same dimension as the individual, and the value of each component is a uniform random number in the interval [-d,1+d], where d is a constant.

6. The method for identifying linear hydrodynamic parameters of underwater vehicles for control law design according to claim 4, characterized in that: In step 5, when performing mutation operation on the individual, Gaussian mutation is used to perform local search to obtain the optimal solution.

Images