Submersible robot shallow diving control method based on improved PSO-ABC algorithm

Abstract

The invention provides a diving robot shallow diving control method based on an improved PSO-ABC algorithm. The method comprises the following steps that a system transfer function of a diving robot in shallow diving at a specific depth is constructed; a PID controller is constructed; fitness functions of Kp, Ki and Kd are designed based on an improved PSO-ABC algorithm; the minimum value of the fitness function is obtained, Kp, Ki and Kd in the corresponding state form a PID controller, water inflow and drainage adjustment of the ballast tank is optimized, and therefore shallow diving of the diving robot at the specific depth is achieved. Based on the improved PSO-ABC algorithm, PID parameter setting has the advantages of being easy to operate, good in robustness, high in convergence speed and the like, and then the control method for shallow diving of the diving robot is provided.

Description

Technical Field

[0001] The present invention relates to the field of robot control, in particular to a shallow diving control method of a diving robot based on an improved PSO-ABC algorithm. Background Art

[0002] The background technology for a shallow-diving control method for a submersible robot based on an improved PSO-ABC algorithm primarily includes basic algorithms such as PSO and ABC, as well as the operating principles of PID controllers and ballast tanks. The particle swarm optimization (PSO) algorithm was inspired by observations of bird flocks foraging, and the artificial bee colony (ABC) algorithm was inspired by the behavior of bees collecting nectar. These algorithms mimic the swarm intelligence found in nature, ultimately solving complex optimization problems. PID control corrects system errors using proportional, integral, and derivative control mechanisms. Its simple structure and robustness make it widely used in various industrial applications. Ballast tanks are a common device for achieving submersible diving, often found in submarines. According to Archimedes' principle, the buoyancy of an object is equal to the weight of the liquid it displaces. When a submarine needs to sink, it uses a compressed air pump to pump water into the ballast tanks, increasing the submarine's weight and causing it to sink. When the submarine needs to surface, it uses a compressed air pump to drain the water from the ballast tanks, reducing the weight and causing it to rise.

[0003] For submersible robots deployed in aquatic environments, the response speed, accuracy, and stability of shallow dives at specific locations require repeated debugging. Directly determining the underwater robot's position through manual observation is not a long-term solution. Therefore, using an improved PSO-ABC algorithm for PID parameter tuning can theoretically enable the submersible robot to hover at any depth underwater without the need for manual observation. This allows for rapid, accurate, and stable shallow dives, offering significant advantages in terms of effectiveness, operational difficulty, and labor costs. Summary of the Invention

[0004] In view of this, an object of the present invention is to provide a children's concentration testing system to at least solve the above problems.

[0005] The technical solution adopted in the present invention is as follows:

[0006] A shallow diving control method for a diving robot based on an improved PSO-ABC algorithm, the method comprising the following steps:

[0007] S1. Construct the system transfer function of the diving robot for shallow diving at a specific depth;

[0008] S2, build PID controller;

[0009] S3. Design of K based on improved PSO-ABC algorithm p , K i , K d The fitness function of

[0010] S4, get the minimum value of fitness function, corresponding to the state K p , K i , K d , forming a PID controller to optimize the system regulation;

[0011] S5. The PID controller responds and dynamically adjusts the water level in the diving robot cabin to achieve shallow diving at a specific depth.

[0012] Furthermore, in step S1, the system transfer function of the submersible robot diving at a specific depth is constructed as follows:

[0013] The submersible robot needs to use the water intake and drainage functions of the ballast tank to achieve shallow diving at a specific depth. The ballast tank can be compared to the classic single-volume water tank model in the industrial field. S represents the bottom area of the ballast tank, and q i , q0 represent the water intake and discharge respectively and are functions of time t, R is the liquid resistance, which represents the resistance of the drain valve, V represents the water storage capacity of the ballast tank, h o represents the steady-state height of the water level, Δh represents the increment of the water level, and u represents the opening of the water inlet valve. According to the dynamic balance relationship of the materials, the following equation is obtained:

[0014]

[0015] Water inlet q i There is the following equation:

[0016] Δq i =k u Δu(2)

[0017] In formula (2), k u Indicates the flow coefficient of the inlet valve,

[0018] According to the kinematic equation, the displacement q0 and the steady-state water level h are expressed as follows:

[0019]

[0020] At the equilibrium point (h o ,q o ), and the relationship between the displacement qo and the fluid resistance R is obtained:

[0021]

[0022] Substituting equations (2), (3), and (4) into equation (1), we can obtain:

[0023]

[0024] Assume time constant T = RS, gain K = k u R, Laplace transform of equation (5) yields the transfer function of the ballast tank:

[0025]

[0026] Furthermore, in step S2, the PID controller is constructed as follows:

[0027] The standard PID control value is proportional to the deviation, the integral of the deviation over time, and the differential of the deviation over time. The corresponding proportional adjustment, integral adjustment, and integral adjustment are performed to obtain the following calculation formula:

[0028]

[0029] Where K p is the proportionality coefficient, K i is the integral coefficient, K d is the differential coefficient, e(t) is the error, which is a function of time t and is defined as follows:

[0030] e(t)=r(t)-y(t)(8)

[0031] Where r(t) is the expected value, y(t) is the actual value,

[0032] The Laplace transform of the PID controller is expressed as follows:

[0033]

[0034] Furthermore, in step S3, K is designed based on the improved PSO-ABC algorithm. p , K i , K d The fitness function is specifically:

[0035] In each iteration, the PSO algorithm is used in the global scope to search for the global optimal value, and the global optimal value and its position at this time are saved. Then the ABC algorithm is used to continue searching for the local optimal value near the global optimal value found by PSO. If the optimal value obtained by ABC is the previously saved global optimal value, the global optimal value will be updated. Otherwise, the global optimal value obtained by PSO will be maintained. At this point, the iteration process ends and enters the next iteration process. After completing all iterative searches, the global optimal value is output, and finally the PID tuning parameters corresponding to the optimal value of the objective function are found.

[0036] PSO is adjusted according to the inertia weight pre-designed by PSO. The inertia weight adjustment relationship is:

[0037]

[0038] Where w is the inertia weight, m is the total number of iterations, and n is the current iteration number (n = 1, 2, ..., m):

[0039] K corresponding to the optimal solution obtained by the improved PSO-ABC algorithm p , K i , K d The PID controller corrects the error in the process of water filling and drainage in the ballast tank, thereby achieving shallow diving of the submersible robot at a specific depth. o represents the steady-state height of the water level, h(t) represents the water level height at a certain moment, and according to formula (8), the error is:

[0040] e(t)=h(t)-h o ;

[0041] According to the error integration criterion, for the inertial link of the ballast tank model, the integrated square error ISE is used to characterize the performance evaluation index of the system. Its fitness function is the water level height h(t) at a certain moment and the water level steady-state height h o The difference between the two, taking the infinite integral of the Euclidean norm, the formula is as follows:

[0042] ISE=∫0 ∞ ||h(t)-h o ||2dt (11).

[0043] Furthermore, in step S4, the minimum value of the fitness function is obtained, corresponding to the state K p , K i , K d , the PID controller is composed to optimize the system regulation as follows:

[0044] Initialize the position and speed of the PSO or ABC individual, the number of algorithm iterations, the inertia weight, and the learning factor parameters, calculate the fitness function value of each individual, obtain the minimum fitness function value and its corresponding PID parameter value, and iterate in the direction of smaller fitness function value.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention proposes a shallow diving control method for a submersible robot based on an improved PSO-ABC algorithm. By using the improved PSO-ABC algorithm for PID parameter tuning, the submersible robot can theoretically hover at any depth underwater. The entire process does not require manual observation, and shallow diving can be achieved quickly, accurately, and stably, which has significant advantages in implementation effect, operation difficulty, and labor cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only preferred embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 This is a schematic diagram of the overall process of the shallow diving control method of the diving robot based on the improved PSO-ABC algorithm.

[0049] Figure 2 This is the closed-loop control schematic diagram

[0050] Figure 3 Ballast tank diagram

[0051] Figure 4 This is the improved PSO-ABC algorithm flow chart

[0052] Figure 5 This is the fitness function convergence effect diagram

[0053] Figure 6 This is the control system simulation diagram

[0054] Figure 7 This is the output and error diagram of the control system before and after correction DETAILED DESCRIPTION

[0055] The principles and features of the present invention are described below with reference to the accompanying drawings. The enumerated embodiments are only used to explain the present invention and are not used to limit the scope of the present invention.

[0056] Reference Figure 1-Figure 7 The present invention provides a shallow diving control method for a diving robot based on an improved PSO-ABC algorithm, the method comprising the following steps:

[0057] S1. Construct the system transfer function of the diving robot for shallow diving at a specific depth;

[0058] S2, build PID controller;

[0059] S3. Design of K based on improved PSO-ABC algorithm p , Ki , K d The fitness function of

[0060] S4, get the minimum value of fitness function, corresponding to the state K p , K i , K d , forming a PID controller to optimize the system regulation;

[0061] S5. The PID controller responds and dynamically adjusts the water level in the diving robot cabin to achieve shallow diving at a specific depth.

[0062] In step S1, the system transfer function of the submersible robot at a specific shallow diving depth is constructed as follows:

[0063] The submersible robot needs to use the water intake and drainage functions of the ballast tank to achieve shallow diving at a specific depth. The ballast tank can be compared to the classic single-volume water tank model in the industrial field. S represents the bottom area of the ballast tank; q i , q0 represent the water intake and discharge respectively and are both functions of time t; R is the liquid resistance, which represents the resistance of the drain valve; V represents the water storage capacity of the ballast tank; h o represents the steady-state height of the water level; Δh represents the increment of the water level; and u represents the opening of the water inlet valve. According to the dynamic balance relationship of the material, the following equation is obtained:

[0064]

[0065] Water inlet q i There is the following equation:

[0066] Δq i =k u Δu (2)

[0067] In formula (2), k u Indicates the flow coefficient of the inlet valve.

[0068] According to the kinematic equation, the displacement q0 and the steady-state water level h are expressed as follows:

[0069]

[0070] At the equilibrium point (h o ,q o ), and the relationship between the displacement qo and the fluid resistance R is obtained:

[0071]

[0072] Substituting equations (2), (3), and (4) into equation (1), we can obtain:

[0073]

[0074] Assume time constant T = RS, gain K = k u R. Performing Laplace transform on equation (5) yields the transfer function of the ballast tank:

[0075]

[0076] In step S2, a PID controller is constructed, specifically:

[0077] PID is an algorithm that performs control through proportional, differential, and integral calculations. The standard PID control value is proportional to the sum of the deviation (the difference between the set value and the actual value of the control system), the integral of the deviation over time, and the differential of the deviation over time. The calculation formula is as follows:

[0078]

[0079] Where K p is the proportionality coefficient, K i is the integral coefficient, K d is the differential coefficient.

[0080] Proportional control is achieved by multiplying the proportional value to quickly adjust the target value. Integral control is achieved by setting an integral value and continuously integrating the deviation based on the deviation to bring the target value closer to the target value. Differential control is achieved by providing a damping function to make the adjustment process smoother.

[0081] e(t) is the error. It is a function of time t. It is defined as follows:

[0082] e(t)=r(t)-y(t)(8)

[0083] Where r(t) is the expected value and y(t) is the actual value.

[0084] The Laplace transform of the PID controller is expressed as follows:

[0085]

[0086] In step S3, K is designed based on the improved PSO-ABC algorithm. p , K i , K d The fitness function is specifically:

[0087] During each iteration of the improved PSO-ABC algorithm, the particle swarm optimization (PSO) algorithm is first used globally to search for the global optimum and save the current global optimum and its location. The artificial bee colony algorithm (ABC) is then used to search for local optima near the global optimum found by the PSO. If the optimal value obtained by the ABC algorithm exceeds the previously saved global optimum, the global optimum is updated. Otherwise, the global optimum obtained by the PSO algorithm is retained. This concludes the current iteration and the next one begins. After all iterative searches are completed, the global optimum is output, ultimately finding the PID tuning parameters that correspond to the optimal value of the objective function.

[0088] For the transfer function model of the submersible robot's suspension process at a specific depth underwater, the algorithm's particle distribution is relatively dispersed at the beginning of the operation, and gradually converges in the later stages. The need to adjust the particle speed and position each time the PSO updates depends on the inertia weight pre-designed by the PSO. The inertia weight decreases as the iteration process progresses, thereby weakening the step size of the particle search process, converting the global search to a fine search, thereby coordinating the further search process near the ABC global optimal value. The inertia weight adjustment relationship is as follows, where w is the inertia weight, m is the total number of iterations, and n is the current iteration number (n = 1, 2, ..., m):

[0089]

[0090] K corresponding to the optimal solution obtained by the improved PSO-ABC algorithm p , K i , K d The PID controller corrects the error in the process of filling and draining the ballast tank, thereby enabling the submersible robot to dive at a specific depth. o represents the steady-state height of the water level, h(t) represents the water level height at a certain moment, and according to formula (8), the error is e(t) = h(t) - h o .

[0091] According to the error integration criterion, for the inertial link of the ballast tank model, the system response oscillation is small, and the integrated square error (ISE) can be used to characterize the performance evaluation index of the system. The fitness function is the water level height h(t) at a certain moment and the water level steady-state height h o The difference between the two, taking the infinite integral of the Euclidean norm, the formula is as follows:

[0092] ISE=∫0 ∞ ||h(t)-h o ||2dt (11)

[0093] In step S4, the minimum value of the fitness function is obtained, corresponding to the state K p, K i , K d , the PID controller is composed to optimize the system regulation as follows:

[0094] Initialize the position and speed of the PSO or ABC individuals, the number of algorithm iterations, the inertia weight, the learning factor, and other parameters. The fitness function value of each individual is calculated, and the minimum fitness function value and the corresponding PID parameter value are obtained. The next iteration is performed in the direction of smaller fitness function values. Ultimately, the PID parameter values converge to a specific position. At this point, the PID controller can quickly, accurately, and stably correct the ballast tank filling and drainage process, and enable the submersible robot to dive shallowly at a specific location.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A shallow diving control method for a diving robot based on an improved PSO-ABC algorithm, characterized in that: The method comprises the following steps: S1. Construct the system transfer function of the diving robot for shallow diving at a specific depth; S2, build PID controller; S3. Design of K based on improved PSO-ABC algorithm p , K i , K d The fitness function of S4, get the minimum value of fitness function, corresponding to the state K p , K i , K d , forming a PID controller to optimize the system regulation; S5. The PID controller responds and dynamically adjusts the water level in the diving robot cabin to achieve shallow diving at a specific depth.

2. The shallow diving control method of a submersible robot based on the improved PSO-ABC algorithm according to claim 1 is characterized in that: In step S1, the system transfer function of the submersible robot diving at a specific depth is constructed as follows: The submersible robot needs to use the water intake and drainage functions of the ballast tank to achieve shallow diving at a specific depth. The ballast tank can be compared to the classic single-volume water tank model in the industrial field. S represents the bottom area of the ballast tank, and q i , q0 represent the water intake and discharge respectively and are functions of time t, R is the liquid resistance, which represents the resistance of the drain valve, V represents the water storage capacity of the ballast tank, h o represents the steady-state height of the water level, Δh represents the increment of the water level, and u represents the opening of the water inlet valve. According to the dynamic balance relationship of the materials, the following equation is obtained: Water inlet q i There is the following equation: Δq i =k u Δu(2) In formula (2), k u Indicates the flow coefficient of the inlet valve, According to the kinematic equation, the displacement q0 and the steady-state water level h are expressed as follows: At the equilibrium point (h o ,q o ), and the relationship between the displacement qo and the fluid resistance R is obtained: Substituting equations (2), (3), and (4) into equation (1), we can obtain: Assume time constant T = RS, gain K = k u R, Laplace transform of equation (5) yields the transfer function of the ballast tank:

3. The shallow diving control method of a submersible robot based on the improved PSO-ABC algorithm according to claim 2 is characterized in that: In step S2, the PID controller is constructed as follows: The standard PID control value is proportional to the deviation, the integral of the deviation over time, and the differential of the deviation over time. The corresponding proportional adjustment, integral adjustment, and integral adjustment are performed to obtain the following calculation formula: Where K p is the proportionality coefficient, K i is the integral coefficient, K d is the differential coefficient, e(t) is the error, which is a function of time t and is defined as follows: e(t)=r(t)-y(t)(8) Where r(t) is the expected value, y(t) is the actual value, The Laplace transform of the PID controller is expressed as follows:

4. The shallow diving control method of a submersible robot based on the improved PSO-ABC algorithm according to claim 3 is characterized in that: In step S3, K is designed based on the improved PSO-ABC algorithm. p , K i , K d The fitness function is specifically: In each iteration, the PSO algorithm is used in the global scope to search for the global optimal value, and the global optimal value and its position at this time are saved. Then the ABC algorithm is used to continue searching for the local optimal value near the global optimal value found by PSO. If the optimal value obtained by ABC is the previously saved global optimal value, the global optimal value will be updated. Otherwise, the global optimal value obtained by PSO will be maintained. At this point, the iteration process ends and enters the next iteration process. After completing all iterative searches, the global optimal value is output, and finally the PID tuning parameters corresponding to the optimal value of the objective function are found. PSO is adjusted according to the inertia weight pre-designed by PSO. The inertia weight adjustment relationship is: Where w is the inertia weight, m is the total number of iterations, and n is the current iteration number (n = 1, 2, ..., m): K corresponding to the optimal solution obtained by the improved PSO-ABC algorithm p , K i , K d The PID controller corrects the error in the process of water filling and drainage in the ballast tank, thereby achieving shallow diving of the submersible robot at a specific depth. o represents the steady-state height of the water level, h(t) represents the water level height at a certain moment, and according to formula (8), the error is: e(t)=h(t)-h o ; According to the error integration criterion, for the inertial link of the ballast tank model, the integrated square error ISE is used to characterize the performance evaluation index of the system. Its fitness function is the water level height h(t) at a certain moment and the water level steady-state height h o The difference between the two, taking the infinite integral of the Euclidean norm, the formula is as follows:

5. The shallow diving control method of a submersible robot based on the improved PSO-ABC algorithm according to claim 4 is characterized in that: In step S4, the minimum value of the fitness function is obtained, corresponding to the state K p , K i , K d , the PID controller is composed to optimize the system regulation as follows: Initialize the position and speed of the PSO or ABC individual, the number of algorithm iterations, the inertia weight, and the learning factor parameters, calculate the fitness function value of each individual, obtain the minimum fitness function value and its corresponding PID parameter value, and iterate in the direction of smaller fitness function value.

Images