A manta ray vehicle heading control method based on fuzzy s face

Abstract

The application relates to a manta ray vehicle heading control method based on fuzzy S surfaces, which comprises the following steps: firstly, acquiring current heading angle information of the manta ray vehicle, and calculating the deviation of the current heading angle and a given heading angle; inputting the heading angle deviation into a heading transition algorithm to obtain a transitioned expected heading angle, taking the transitioned expected heading as the expected heading angle at the current time; recalculating the deviation and the change rate of the deviation of the heading angle; inputting the current heading deviation and the change rate of the deviation of the manta ray vehicle into a fuzzy controller to output the change amount Delta k1 and Delta k2 of the control parameters k1 and k2 of the S surface controller; online correcting the corresponding control parameters k1 and k2 in the S surface controller by using Delta k1 and Delta k2 respectively to obtain an S surface control algorithm based on the transition process; and controlling the pectoral fin movement of the manta ray vehicle by using the control algorithm, so that the manta ray vehicle can track the expected heading.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of underwater vehicle control, and relates to a manta ray vehicle control method based on fuzzy S surface. BACKGROUND

[0002] The manta ray vehicle is a bionic vehicle that uses MPF (central / counter fin mode) mode for propulsion. Unlike the traditional underwater vehicle that uses propeller propulsion, the manta ray vehicle generates propulsion force and torque through the swing of the pectoral fin. Compared with the traditional underwater vehicle, the manta ray vehicle has many advantages such as high propulsion efficiency, good maneuverability, low noise and the like.

[0003] As a new type of bionic underwater vehicle, the heading control of the manta ray vehicle is a basic control task and is the basis for realizing more complex tasks. At present, the methods for controlling underwater vehicles include PID control and fuzzy control, but the PID control method has the problem of difficult parameter setting, and the fuzzy control needs to rely on human experience. The S surface control algorithm is widely used in the control system of underwater vehicles because it does not depend on the accurate mathematical model of the controlled object, has the idea of fuzzy control and the form of classical PD control. However, it is necessary to realize online correction of the parameters of the S surface controller to further improve the control effect of the bionic vehicle because a single control parameter cannot meet the complex working conditions of the bionic vehicle in underwater work. SUMMARY

[0004] Technical problems to be solved

[0005] In order to avoid the shortcomings of the prior art, the application provides a manta ray vehicle control method based on fuzzy S surface. The control method can realize online correction of the parameters of the S surface controller, improve the performance of the S surface controller, increase the anti-interference ability of the manta ray vehicle, realize real-time adjustment of the heading of the manta ray vehicle, introduce a transition process to buffer the influence of the change of the heading angle error on the control output, realize stable control of the manta ray vehicle during sharp turning, and improve the control precision during heading control.

[0006] Technical scheme

[0007] A manta ray vehicle control method based on fuzzy S surface, characterized in that the steps are as follows:

[0008] Step 1: Real-time acquisition of the heading angle ψ(t) of the manta ray vehicle through the attitude sensor carried by the manta ray vehicle to obtain the current heading angle of the manta ray vehicle.

[0009] Difference between the expected heading angle ψ0 and the current heading angle ψ(t) to calculate the heading angle error Δψ; and calculation of the heading angle error change rate Δψ'.

[0010] Step 2: The target heading angle is arranged with a transition process for the deviation value of the current heading angle, and the target heading angle during the transition is calculated by the transition process function, and the current target heading angle is calculated for the left turn and right turn of the manta ray vehicle respectively;

[0011] When the heading angle of the manta ray vehicle suddenly increases, the transition heading angle is calculated as:

[0012]

[0013] When the heading angle of the manta ray vehicle suddenly decreases, the transition heading angle is calculated as:

[0014]

[0015] In the formula, ψ0' is the transition expected heading angle, ψ t0 is the current heading angle of the manta ray vehicle at the beginning of the turn, and ψ1 is the transition heading angle threshold value;

[0016] Step 3: After obtaining the transition heading angle, the current heading angle deviation and the deviation change rate are recalculated, and the current heading deviation e and the deviation change rate are obtained by normalizing the heading angle deviation and the deviation change rate

[0017] Step 4: The current heading deviation e and the deviation change rate are input into the fuzzy controller, and the correction amount Δk1 and Δk2 of the controller parameters are output;

[0018] Step 5: According to the input heading deviation e and heading deviation change rate , the fuzzy value of the S-surface control parameter is obtained by querying the fuzzy control rule table;

[0019] Step 6: The fuzzy value of the control parameter in the S-surface control algorithm is clarified by combining the Mamdani reasoning algorithm and the barycentric method, and the clarified S-surface control parameter correction amount Δk1 and Δk2 are obtained;

[0020] Step 7: The S-surface control algorithm corresponding control parameters k1 and k2 are respectively adjusted by Δk1 and Δk2:

[0021]

[0022] In the formula, k1' is the corrected control parameter k1, k10 is the initial value of the control parameter k1, Δk1 is the correction amount of the control parameter k1, k2' is the corrected control parameter k2, k20 is the initial value of the control parameter k2, and Δk2 is the correction amount of the control parameter k2;

[0023] Step 8: The output of the S surface is multiplied by a scale factor to map to the executable region of the controller, and the executable amount of the manta ray vehicle pectoral fin is obtained, and the manta ray vehicle is controlled.

[0024] The S surface control algorithm is Wherein, is the normalized deviation rate; k1 is the control parameter corresponding to the deviation, and k2 is the control parameter corresponding to the deviation rate; f is the control output.

[0025] The value range of the control output f is [-1, 1].

[0026] In the S surface parameter setting process, the normalized heading angle error e and the heading angle error rate are input into the fuzzy controller, and the correction amount Δk1 and Δk2 of the control parameter are output through fuzzy reasoning; wherein the fuzzy sets of input and output are defined as {NB, NM, NS, ZE, PS, PM, PB}, wherein NB is negative big, NM is negative medium, NS is negative small, ZE is zero, PS is positive small, PM is positive medium, and PB is positive big.

[0027] The Mamdani reasoning algorithm and the barycenter method use 7x7 fuzzy reasoning rules in the S surface control algorithm, and the fuzzy language variables of the S surface control parameters are obtained through fuzzy control rule table query reasoning; finally, defuzzification is carried out, and the barycenter method is used for defuzzification to obtain the clear S surface parameter correction amount Δk1 and Δk2;

[0028] The fuzzy rule table

[0029]

[0030]

[0031] The executable amount f of the manta ray vehicle pectoral fin u = K u × f, wherein f u is the pectoral fin movement parameter of the manta ray vehicle, the pectoral fin swing amplitude or the phase difference between the fin strips, f is the output of the controller, and K u is the control gain.

[0032] The control gain K u is determined by the maximum movable range of the pectoral fin mechanical structure.

[0033] Beneficial effects

[0034] The application provides a manta ray vehicle control method based on a fuzzy S surface, which comprises the following steps: obtaining current heading angle information of the manta ray vehicle, and calculating a deviation of the current heading angle from a given heading angle; inputting the heading angle deviation into a heading transition algorithm to obtain a transitioned expected heading, taking the transitioned expected heading as the expected heading angle at the current time; recalculating the heading angle deviation and a deviation change rate; inputting the current heading deviation and the deviation change rate of the manta ray vehicle into a fuzzy controller to output change amounts Δk1 and Δk2 of control parameters k1 and k2 of the S surface controller; performing online correction on the corresponding control parameters k1 and k2 in the S surface controller by using Δk1 and Δk2 respectively to obtain an S surface control algorithm based on the transition process; and controlling the pectoral fin movement of the manta ray vehicle by using the control algorithm to make the manta ray vehicle track the expected heading.

[0035] Compared with the existing bionic underwater vehicle control method, the application has the advantages that:

[0036] 1. The S surface control method is used to design a heading controller, the controller has a simple structure, is easy to implement, and has less dependence on human experience compared with the fuzzy control method.

[0037] 2. The fuzzy control method is used to realize online correction of the control parameters of the S surface controller, and the adaptability of the vehicle to different working conditions is improved.

[0038] 3. The heading transition process is arranged for the case that the target heading angle changes suddenly, the influence of the change of the heading angle error on the control output is buffered, the stable control of the vehicle heading during the sharp turning is realized, and the control precision is improved. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 The flow chart of the application

[0040] Figure 2 The control block diagram of the application

[0041] Figure 3 The experimental result diagram of the method of the application

[0042] Figure 4 The experimental result comparison diagram of the method of the application and the classical S surface control method DETAILED DESCRIPTION

[0043] The application will be further described in combination with the embodiments and the drawings:

[0044] The application provides a manta ray vehicle control method based on a fuzzy S surface, which comprises the following steps: obtaining current heading angle information of the manta ray vehicle, and calculating a deviation of the current heading angle from a given heading angle; inputting the heading angle deviation into a heading transition algorithm to obtain a transitioned expected heading, taking the transitioned expected heading as the expected heading angle at the current time; recalculating the heading angle deviation and a deviation change rate; inputting the current heading deviation and the deviation change rate of the manta ray vehicle into a fuzzy controller to output change amounts Δk1 and Δk2 of control parameters k1 and k2 of the S surface controller; performing online correction on the corresponding control parameters k1 and k2 in the S surface controller by using Δk1 and Δk2 respectively to obtain an S surface control algorithm based on the transition process; and controlling the pectoral fin movement of the manta ray vehicle by using the control algorithm to make the manta ray vehicle track the expected heading.

[0045] The manta ray vehicle collects the heading angle information of the manta ray vehicle in real time through the attitude sensor carried by the manta ray vehicle, and obtains the current heading angle of the manta ray vehicle.

[0046] The heading angle error is calculated according to formula (1):

[0047] Given the desired heading angle, the deviation of the desired heading angle and the current heading angle is calculated, and the calculation method is shown in formula (1):

[0048] Δψ=ψ(t)-ψ0 (1)

[0049] In the formula, Δψ is the current heading angle error, ψ(t) is the current heading angle of the manta ray vehicle, and ψ0 is the desired heading angle.

[0050] The heading angle error rate is calculated according to formula (2):

[0051] Δψ'=ψ(t)-ψ(t-1) (2)

[0052] In the formula, Δψ' is the heading angle error rate, ψ(t) is the current heading angle of the manta ray vehicle, and ψ(t-1) is the heading angle of the manta ray vehicle at the last time.

[0053] The target heading angle is arranged for the deviation value of the current heading angle, and the target heading angle during the transition period is calculated through the transition process function as the current target heading angle. The left turn and right turn of the manta ray vehicle are calculated respectively, and the calculation method is as follows:

[0054] When the heading angle of the manta ray vehicle suddenly increases, the calculation method of the transition heading angle is shown in formula (3):

[0055]

[0056] When the heading angle of the manta ray vehicle suddenly decreases, the calculation method of the transition heading angle is shown in formula (4):

[0057]

[0058] In the formula, ψ0' is the transition desired heading angle, ψ t0 is the current heading angle of the manta ray vehicle at the beginning of the turn, and ψ1 is the transition heading angle threshold.

[0059] After obtaining the transition heading angle, the current heading angle deviation and the deviation rate and their normalized values are recalculated according to formulas (1), (2), (3) and (4).

[0060] The normalized current heading deviation e of the manta ray vehicle and its deviation rate The input fuzzy controller outputs the correction amount Δk1 and Δk2 of the controller parameters; according to the input heading deviation e and the change rate of the heading deviation The fuzzy value of the S-surface control parameter is obtained by querying the fuzzy control rule table; finally, the fuzzy value of the control parameter in the S-surface control algorithm is clarified by combining the Mamdani reasoning algorithm and the barycenter method to obtain the clarified S-surface control parameter correction amount Δk1 and Δk2.

[0061] Δk1 and Δk2 are used to respectively correct the corresponding control parameters k1 and k2 in the S-surface control algorithm according to formula (5).

[0062]

[0063] In the formula, k1' is the corrected control parameter k1, k10 is the initial value of the control parameter k1, Δk1 is the correction amount of the control parameter k1, k2' is the corrected control parameter k2, k20 is the initial value of the control parameter k2, and Δk2 is the correction amount of the control parameter k2.

[0064] The S-surface algorithm used in the method is shown in formula (6):

[0065]

[0066] wherein, is the normalized change rate of the deviation; k1 is the control parameter corresponding to the deviation, and k2 is the control parameter corresponding to the change rate of the deviation. f is the control output, and its value range is [-1, 1]. Therefore, the S-surface output needs to be mapped to the executable interval of the actuator through a gain factor in the control process. Specific embodiments

[0068] The manta ray vehicle includes a power module, a processor module, a sensor module, and an actuator module, and the overall block diagram of the system is as shown in Figure 1 The attitude sensor in the sensor module is used to collect the heading angle information of the manta ray vehicle in real time, and the collected data is used as the current heading angle of the manta ray vehicle.

[0069] Figure 2It is a kind of based on transition value function fuzzy S surface manta ray vehicle heading control principle diagram proposed in the application, including input signal, transition process module, S surface adaptive fuzzy controller module, conventional s surface controller module, controlled object manta ray vehicle, attitude sensor, system output quantity;Input signal is ultimately through S surface controller module, acts on the controlled object manta ray vehicle, and the heading angle signal collected by the attitude sensor is compared with the input target heading angle signal, and the error amount of signal is input as input variable into the transition process module, and the expected heading angle after transition is calculated, and the new calculated heading angle deviation is input as input variable into the S surface adaptive fuzzy controller module, and acts on the conventional S surface controller module, to realize the online setting of S surface controller parameters and the closed-loop control to the manta ray vehicle heading.

[0070] The application proposes a kind of based on transition process fuzzy S surface manta ray vehicle heading control method, and the specific steps are as follows:

[0071] Manta ray vehicle is collected by the attitude sensor carried by itself in real time the heading angle information of manta ray vehicle, obtains the current heading angle of manta ray vehicle.

[0072] Desired heading angle is given by control system, the difference between desired heading angle and current heading angle is calculated by the current heading angle information collected in real time in step 1, and the input current heading angle error Δ of S surface controller and heading angle error rate Δ are obtained.

[0073] Wherein, the calculation method of heading angle error is as shown in formula (7):

[0074] Δψ=ψ(t)-ψ0 (7)

[0075] In the formula, Δψ is current heading angle error, ψ (t) is the current heading angle of manta ray vehicle, and ψ0 is desired heading angle.

[0076] The calculation method of heading angle error rate is as shown in formula (8):

[0077] Δψ'=ψ(t)-ψ(t-1) (8)

[0078] In the formula, Δψ' is the heading angle error rate, ψ (t) is the current heading angle of manta ray vehicle, and ψ (t-1) is the heading angle of manta ray vehicle at last time.

[0079] The current heading angle error Δ and the heading angle error rate Δ are normalized to obtain the input of fuzzy controller, and the specific calculation mode is as follows:

[0080] The normalization calculation method of heading angle error is as shown in formula (9):

[0081] e = Δψ / Δψ max (9)

[0082] In the formula, e is the normalized heading angle error, Δψ is the heading angle error, and Δψ max This represents the maximum value of the heading angle error.

[0083] The normalized calculation method for the rate of change of heading angle error is shown in Equation (10):

[0084]

[0085] In the formula, Let Δψ' be the normalized rate of change of heading angle error. max ' is the maximum value of the rate of change of heading angle error.

[0086] The transition process is designed primarily to address the stability control issue when the manta ray simulator's heading angle changes abruptly. Therefore, to achieve the desired heading angle adjustment during the manta ray simulator's turning process, the transition heading angle is calculated separately for both left and right turns. The specific calculation method is shown below:

[0087] When the heading angle of the manta ray-inspired vehicle suddenly increases, the transition heading angle is calculated as shown in equation (11).

[0088]

[0089] When the heading angle of the manta ray-inspired vehicle suddenly decreases, the transition heading angle is calculated as shown in equation (12):

[0090]

[0091] In the formula, ψ0' is the expected heading angle due to the transition, ψ t0 ψ1 represents the current heading angle of the manta ray-like vehicle at the moment of starting the turn, and ψ1 represents the threshold for the transition heading angle.

[0092] After obtaining the transition heading angle, the heading angle deviation and the rate of change of deviation and their normalized values ​​at the current time are recalculated according to formulas (7), (8), (9), and (10).

[0093] S-plane control parameters tuning: During the S-plane parameter tuning process, the normalized heading angle error e and the heading angle error change rate are... The correction amount of the control parameter Δk1 and Δk2 is outputted through fuzzy inference by inputting into the fuzzy controller; wherein the fuzzy sets of input and output are defined as {NB, NM, NS, ZE, PS, PM, PB}, wherein NB is negative big, NM is negative medium, NS is negative small, ZE is zero, PS is positive small, PM is positive medium, and PB is positive big; the membership functions of input and output are all selected as triangular functions; the fuzzy inference rule of 7x7 is adopted, the fuzzy language variable of the S surface control parameter is obtained through fuzzy control rule table inquiry inference, wherein the fuzzy rule table of k1 and k2 correction is shown in Table 1 and Table 2; finally, defuzzification is carried out, the defuzzification is carried out by using the barycentric method, and the clear S surface parameter correction amount Δk1 and Δk2 are obtained.

[0094] Table 1 Parameter k1 correction rule

[0095]

[0096] Table 2 Parameter k2 correction rule

[0097]

[0098] Δk1 and Δk2 are respectively used to correct the corresponding control parameters k1 and k2 in the S surface control algorithm in real time according to formula (13).

[0099]

[0100] In the formula, k1' is the corrected control parameter k1, k10 is the initial value of the control parameter k1, Δk1 is the correction amount of the control parameter k1, k2' is the corrected control parameter k2, k20 is the initial value of the control parameter k2, and Δk2 is the correction amount of the control parameter k2.

[0101] The S surface control algorithm used in the application is shown in formula (14):

[0102]

[0103] Wherein, is the normalized deviation rate; k1 is the control parameter corresponding to the deviation, and k2 is the control parameter corresponding to the deviation rate. f is the control output, and the value range is [-1, 1]. Therefore, in order to obtain the executable amount of the manta ray vehicle pectoral fin, the output of the S surface needs to be multiplied by a proportional factor to map to the executable area of the controller, and the specific formula is shown in formula (15):

[0104] f u =K u ×f(15)

[0105] In the formula, f uFor the motion parameters of the manta ray vehicle's pectoral fin, the pectoral fin swing amplitude or the phase difference between the fin strips, f is the output of the controller, K u is the control gain, the size of which is determined by the maximum movable range of the mechanical structure of the pectoral fin.

[0106] The mapped controller output quantity is input to the actuator driving the pectoral fin, the actuator makes corresponding action, the pectoral fin swing makes the heading angle of the manta ray vehicle change, and moves to the desired heading.

Claims

1. A manta ray vehicle control method based on fuzzy S surface, characterized in that The steps are as follows: Step 1: Real-time collection of the heading angle ψ(t) of the manta ray vehicle through the attitude sensor carried by the manta ray vehicle itself, to obtain the current heading angle of the manta ray vehicle; Difference with the expected heading angle ψ0, to calculate the heading angle error Δψ; Calculate the heading angle error rate Δψ'; Step 2: Arrange a transition process for the target heading angle according to the deviation value of the current heading angle, calculate the target heading angle during the transition period through the transition process function, and calculate the current target heading angle for the left turn and right turn of the manta ray vehicle respectively; When the heading angle of the manta ray vehicle suddenly increases, the calculation of the transition heading angle is: When the heading angle of the manta ray vehicle suddenly decreases, the calculation of the transition heading angle is: wherein ψ0' is a desired heading angle of the manta ray vehicle during the transition period, ψ0 is a current heading angle of the manta ray vehicle at the start of the transition period, ψ1 is a threshold value for the transition heading angle, and ψ is a current heading angle of the manta ray vehicle at the time t. t0 wherein ψ0' is a desired heading angle of the manta ray vehicle during the transition period, ψ0 is a current heading angle of the manta ray vehicle at Step 3: After the transition heading angle is obtained, the heading angle deviation and the deviation rate of change at the current time are recalculated, and the heading angle deviation and the deviation rate of change are normalized to obtain the current heading deviation e and the deviation rate of change thereof Step 4: the current course deviation e and its deviation change rate inputting the fuzzy controller, and outputting the correction amounts Δki and Δk2 of the controller parameters; Step 5: Based on the input heading deviation e and heading deviation rate of change Obtaining the fuzzy value of the S-surface control parameter from the query fuzzy control rule table; Step 6: Clear the fuzzy value of the control parameter in the S surface control algorithm by combining the Mamdani reasoning algorithm and the barycentric method, to obtain the clear S surface control parameter correction amount Δk1 and Δk2; Step 7: Real-time adjustment of the corresponding control parameters k1 and k2 in the S surface control algorithm by Δk1 and Δk2 respectively: In the formula, k1' is the corrected control parameter k1, k10 is the initial value of the control parameter k1, Δk1 is the correction amount of the control parameter k1, k2' is the corrected control parameter k2, k20 is the initial value of the control parameter k2, and Δk2 is the correction amount of the control parameter k2. Step 8: Multiply the output of the s surface by a proportional factor to map to the executable area of the controller, to obtain the executable amount of the pectoral fin of the manta ray vehicle, and control the manta ray vehicle.

2. The method according to claim 1, wherein: The S surface control algorithm is wherein, is the normalized deviation change rate; k1 is a control parameter corresponding to the deviation, k2 is a control parameter corresponding to the deviation change rate; and f is a control output.

3. The manta ray-based vehicle control method based on fuzzy S surface according to claim 2, characterized in that: The value range of the control output f is [-1, 1].

4. The method according to claim 1, wherein: In the S surface parameter setting process, the normalized heading angle error e and the heading angle error change rate By inputting into the fuzzy controller, the correction amount Δk1 and Δk2 of the control parameter are outputted through fuzzy inference; wherein the fuzzy sets of input and output are defined as {NB, NM, NS, ZE, PS, PM, PB}, wherein, NB is negative big, NM is negative medium, NS is negative small, ZE is zero, PS is positive small, PM is positive medium, and PB is positive big.

5. The method of claim 1, wherein: The Mamdani reasoning algorithm and the barycentric method in the S surface control algorithm adopt 7x7 fuzzy reasoning rules, obtain the fuzzy language variable of the S surface control parameter through the fuzzy control rule table query reasoning, and finally de-fuzzify, adopt the barycentric method to de-fuzzify, to obtain the clear S surface parameter correction amount Δk1 and Δk2; The fuzzy rule table 6. The method of claim 1, wherein: The executable quantity f of the manta ray vehicle pectoral fin u = K u × f, in the formula, f u is the pectoral fin motion parameter of the manta ray vehicle, the pectoral fin swing amplitude or the phase difference between fin strips, f is the output of the controller, K u is the control gain.

7. The manta ray-based Sf surface control method according to claim 6, wherein: The control gain K u The size is determined by the maximum movable range of the mechanical structure of the pectoral fin.

Images