An ant colony algorithm-based high-order iterative learning control method, device and computer equipment applied to a single-link manipulator non-uniform interval

Abstract

The application relates to a high-order iterative learning control method and device based on an ant colony algorithm applied to a non-uniform interval of a single-link manipulator and computer equipment, the method comprising the following steps: optimizing a pre-constructed target single-link manipulator model by using an ant colony algorithm to obtain optimal control gain; designing a sampling error function of the target single-link manipulator system; designing an iterative learning control law of the target single-link manipulator model based on the sampling error function and the optimal control gain; controlling the target single-link manipulator model to perform high-order iterative learning, and updating a control signal of next iteration of the target single-link manipulator model by using the iterative learning control law until the sampling error function converges, and the iteration is stopped, so that the current control signal of the target single-link manipulator system is obtained. The application has the effects of faster convergence speed and higher control precision.

Description

Technical Field

[0001] This application relates to the field of single-link manipulator control technology, and in particular to a high-order iterative learning control method, device and computer equipment based on ant colony algorithm for non-uniform regions of single-link manipulators. Background Technology

[0002] Single-link robotic arms are very common in industry and daily life, such as industrial robotic arms used in manufacturing and claw machines frequently seen in daily life. The working mode of a single-link robotic arm is usually to continuously perform repetitive operations within the same operating range, such as grasping and transporting. However, in actual production processes, some unavoidable emergencies often occur, such as items falling or equipment malfunctions, which causes the operating range of the robotic arm system to change randomly.

[0003] As industrial production processes become increasingly complex, it is generally difficult to establish accurate system models for trajectory control problems involving nonlinearity and complexity. Single-link manipulator control methods based on iterative learning control algorithms can effectively solve the problem of inaccurate trajectory tracking caused by inaccurate system models in traditional control methods. Furthermore, these methods require less prior knowledge and computational resources, enabling them to handle highly uncertain dynamic systems in a very simple way. However, existing single-link manipulator control methods based on iterative learning algorithms often fall into a vicious cycle of iteration due to a lack of clear control objectives and requirements, resulting in slow convergence speeds.

[0004] Regarding the aforementioned technologies, the inventors have discovered that existing single-link manipulator control methods based on iterative learning control algorithms suffer from slow convergence speed and poor control performance for single-link manipulators. Summary of the Invention

[0005] To accelerate the convergence speed of single-link manipulator control methods and improve the control effect of single-link manipulators, this application provides a high-order iterative learning control method, device, and computer equipment based on ant colony algorithm for non-uniform regions of single-link manipulators.

[0006] In the first aspect, this application provides a high-order iterative learning control method based on ant colony algorithm for non-uniform intervals of a single-link manipulator.

[0007] This application is achieved through the following technical solution:

[0008] A high-order iterative learning control method based on ant colony algorithm for non-uniform regions of a single-link manipulator includes the following steps:

[0009] Obtain the current angular position and current velocity of the target single-link manipulator system, and construct a target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia, and expected working time of the target single-link manipulator system.

[0010] The target single-link manipulator model is optimized using the ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model;

[0011] The sampling period of the target single-link manipulator system is preset to determine the current desired output trajectory and the current control signal, and the current control signal is input into the target single-link manipulator system to obtain the current output signal.

[0012] Based on the current desired output trajectory and the current output signal, design the sampling error function of the target single-link manipulator system;

[0013] Based on the sampling error function and the optimal control gain, an iterative learning control law for the target single-link manipulator model is designed.

[0014] The target single-link manipulator model is controlled to perform high-order iterative learning, and the control signal for the next iteration of the target single-link manipulator model is updated using the iterative learning control law until the sampling error function converges and the iteration stops. The current control signal is then obtained to control the target single-link manipulator system.

[0015] In a preferred embodiment, this application can be further configured such that: the step of optimizing the target single-link manipulator model using an ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model includes...

[0016] Initialize the target parameters of the ant colony algorithm;

[0017] Based on the target single-link manipulator model, the current angular position of the target single-link manipulator is randomly set, the fitness function for obtaining pheromones is calculated, the initial pheromones are obtained, and the state transition probability is calculated.

[0018] The initial pheromone and the state transition probability are input into the fitness function to iteratively optimize the target single-link manipulator model and obtain the updated pheromone.

[0019] Determine whether the number of iterations for the target single-link manipulator model has reached the maximum number of iterations;

[0020] If the number of iterations for the target single-link manipulator model reaches the maximum number of iterations, the current control gain of the target single-link manipulator model is taken as the optimal control gain and output.

[0021] In a preferred embodiment, this application can be further configured such that, when designing the sampling error function of the target single-link manipulator system based on the current desired output trajectory and the current output signal, the following formula is used.

[0022]

[0023] E{λ k (n)}=q k (n);

[0024]

[0025]

[0026] In the formula, This represents the sampling error value of the target single-link manipulator system at time n during the k-th iteration; The target single-link manipulator system represents the expected output trajectory sample value at time n; λ represents the sampled output signal value of the target single-link manipulator system at time n; k qk(n) represents a Bernoulli random variable at time n, taking values ​​of 0 and 1; qk(n) represents a Bernoulli random variable λ. k (n) represents the expected value at time n; ek(n) represents the error value of the target single-link manipulator system in the k-th iteration at time n; T S T represents the sampling period of the target single-link manipulator system; d T represents the expected working time of the manipulator in the target single-link manipulator system. K This represents the current moment corresponding to the control signal of the target single-link manipulator system.

[0027] In a preferred embodiment, this application can be further configured such that, when designing the iterative learning control law for the target single-link manipulator model based on the sampling error function and the optimal control gain, the following formula is used.

[0028]

[0029]

[0030] |W v -L v C(n+1)B(n)|≤γ v ;

[0031]

[0032]

[0033] C = [0 1];

[0034] In the formula, V represents the sampled value of the control signal input to the target single-link manipulator model at time n during the (k+1)th iteration; V represents the higher order, ranging from 1, 2, ..., N; W v L represents the optimal control gain applied to the control input by the ant colony algorithm. v This represents the optimal control gain of the ant colony algorithm applied to the sampling error function; This represents the sampled value of the control signal input to the target single-link manipulator model at time n during the (k-V+1)th iteration. T represents the sampling error value of the target single-link manipulator system at time n in the (k-V+1)th iteration; S T represents the sampling period of the target single-link manipulator system; d J represents the expected working time of the manipulator in the target single-link manipulator system; J represents the moment of inertia of the relative joints of the target single-link manipulator system.

[0035] In a preferred embodiment, this application can be further configured such that, when updating the control signal for the next iteration of the target single-link manipulator model using the iterative learning control law, the following formula is employed.

[0036]

[0037] In the formula, u k+1 (t) represents the control signal of the target single-link manipulator model at time t for the (k+1)th iteration; The sampled value of the control signal for the (k+1)th iteration of the target single-link manipulator model at time t; T S T represents the sampling period of the target single-link manipulator system; d The expected working time of the manipulator in the target single-link manipulator system.

[0038] In a preferred embodiment, this application can be further configured such that the expression for the target single-link manipulator model is as follows:

[0039]

[0040] In the formula, The sampled value of the angle position of the single-link manipulator during the k-th iteration at time t; x is the sampled value of the velocity of the single-link manipulator during the k-th iteration at time t; 2,k (t) represents the velocity of the single-link manipulator during the k-th iteration at time t; x 1,k(t) represents the angular position of the single-link manipulator at time t during the k-th iteration; J represents the relative moment of inertia of the target single-link manipulator system; u k (t) represents the control signal input to the target single-link manipulator system at time t; m0 represents the mass of the single-link manipulator; M0 is the tip load of the single-link manipulator; g is the acceleration due to gravity; l represents the length of the single-link manipulator; y k (t) represents the output signal of the target single-link manipulator system at time t; T K T represents the current time corresponding to the control signal of the target single-link manipulator system; d The expected working time of the manipulator in the target single-link manipulator system.

[0041] In a preferred embodiment, this application can be further configured such that: when the sampling error function converges, it converges to within the tolerance error, and the expression of the error index function of the tolerance error is as follows.

[0042]

[0043] In the formula, ME k The value of the function representing the error index; y d (n) represents the desired output trajectory of the target single-link manipulator system at time n; y k (n) represents the output signal of the target single-link manipulator system at time n; T d T represents the expected working time of the manipulator in the target single-link manipulator system. S The sampling period represents the target single-link manipulator system.

[0044] In a preferred embodiment, this application can be further configured to include the following steps when presetting the sampling period of the target single-link manipulator system to determine the current desired output trajectory:

[0045] The expression for the output signal of the target single-link manipulator system at time t is as follows:

[0046]

[0047] In the formula, y d (t) represents the desired output signal of the target single-link manipulator system at time t;

[0048] According to the preset sampling period T S For y d (t) is sampled to obtain the expected output sequence y of the target single-link manipulator system at time t. d (t·T S ), that is, the desired output trajectory of the target single-link manipulator system at time t.

[0049] Secondly, this application provides a high-order iterative learning control device based on ant colony algorithm for use in non-uniform regions of a single-link manipulator.

[0050] This application is achieved through the following technical solution:

[0051] A high-order iterative learning control device based on ant colony algorithm for non-uniform regions of a single-link manipulator, comprising:

[0052] The system model module is used to obtain the current angular position and current velocity of the target single-link manipulator system, and to construct the target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia and expected working time of the target single-link manipulator system.

[0053] The control gain optimization module is used to optimize the target single-link manipulator model using the ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model.

[0054] The system output module is used to preset the sampling period of the target single-link manipulator system to determine the current desired output trajectory and the current control signal, and input the current control signal into the target single-link manipulator system to obtain the current output signal;

[0055] The system error module is used to design the sampling error function of the target single-link manipulator system based on the current expected output trajectory and the current output signal.

[0056] The control law module is used to design the iterative learning control law of the target single-link manipulator model based on the sampling error function and the optimal control gain.

[0057] An iterative module is used to control the target single-link manipulator model to perform high-order iterative learning, and to update the control signal of the target single-link manipulator model for the next iteration using the iterative learning control law, until the sampling error function converges and the iteration stops, and the current control signal is obtained to control the target single-link manipulator system.

[0058] Thirdly, this application provides a computer device.

[0059] This application is achieved through the following technical solution:

[0060] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of any of the above-described high-order iterative learning control methods based on ant colony algorithm applied to the non-uniform region of a single-link manipulator.

[0061] In summary, compared with the prior art, the beneficial effects of the technical solution provided in this application include at least the following:

[0062] A target single-link manipulator model is constructed to represent the actual single-link manipulator system. The ant colony algorithm is used to optimize the target single-link manipulator model, resulting in better control gain in the iterative learning control method for non-uniform intervals. The sampling period of the target single-link manipulator system is preset to determine the current desired output trajectory and the current control signal. The current control signal is then input into the target single-link manipulator system to obtain the current output signal. Based on the current desired output trajectory and the current output signal, a sampling error function is designed to correct the control signal error during the iteration process, thereby improving the control accuracy of the target single-link manipulator system. Based on the sampling error function and the optimal control gain, a target single-link manipulator is designed... The iterative learning control law of the robotic arm model controls the target single-link robotic arm model to perform high-order iterative learning, and uses the iterative learning control law to update the control signal of the target single-link robotic arm model for the next iteration, so as to help the target single-link robotic arm model learn the optimal control signal and control the target single-link robotic arm system to obtain better system output. Compared with the first-order iterative learning control algorithm, the high-order iterative learning control algorithm (HOILC) of this application updates the model's next input by learning the tracking information of the previous few iterations. It has a faster convergence speed than the traditional high-order iterative learning control method, so the tracking performance is better than the first-order iterative learning control algorithm. This application has higher control accuracy for the single-link robotic arm, with excellent performance indicators and wider applicability. Attached Figure Description

[0063] Figure 1 This is a flowchart illustrating a high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator, as an exemplary embodiment of this application.

[0064] Figure 2 This is a schematic diagram illustrating the ant colony algorithm optimization principle of a high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator, as another exemplary embodiment of this application.

[0065] Figure 3This diagram illustrates the relationship between the system output and the desired output trajectory in the 5th, 10th, and 30th iterations of a high-order iterative learning control method based on ant colony algorithm for non-uniform regions of a single-link manipulator, as provided in another exemplary embodiment of this application.

[0066] Figure 4 This diagram illustrates the relationship between the number of iterations and the iterative tracking error of a high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator, as provided in an exemplary embodiment of this application, and a traditional high-order iterative learning control method.

[0067] Figure 5 This application provides a structural block diagram of a high-order iterative learning control device based on ant colony algorithm for use in non-uniform regions of a single-link manipulator, as an exemplary embodiment of the present application. Detailed Implementation

[0068] This specific embodiment is merely an explanation of this application and is not intended to limit it. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application.

[0069] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0070] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0071] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0072] Reference Figure 1 This application provides a high-order iterative learning control method based on ant colony algorithm for non-uniform intervals of a single-link manipulator. The main steps of the method are described below.

[0073] S1: Obtain the current angular position and current velocity of the target single-link manipulator system, and construct a target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia, and expected working time of the target single-link manipulator system.

[0074] S2: The target single-link manipulator model is optimized using the ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model;

[0075] S3: Preset the sampling period of the target single-link manipulator system to determine the current desired output trajectory and the current control signal, and input the current control signal into the target single-link manipulator system to obtain the current output signal;

[0076] S4: Based on the current desired output trajectory and the current output signal, design the sampling error function of the target single-link manipulator system;

[0077] S5: Based on the sampling error function and the optimal control gain, design the iterative learning control law for the target single-link manipulator model;

[0078] S6: Control the target single-link manipulator model to perform high-order iterative learning, and use the iterative learning control law to update the control signal of the target single-link manipulator model for the next iteration until the sampling error function converges and the iteration stops, so as to obtain the current control signal for controlling the target single-link manipulator system.

[0079] Specifically, the expression for a target single-link manipulator model is as follows:

[0080]

[0081] In the formula, The sampled value of the angle position of the single-link manipulator during the k-th iteration at time t; x is the sampled value of the velocity of the single-link manipulator during the k-th iteration at time t; 2,k (t) represents the velocity of the single-link manipulator during the k-th iteration at time t; x 1,k (t) represents the angular position of the single-link manipulator at time t during the k-th iteration; J represents the relative moment of inertia of the target single-link manipulator system; u k (t) represents the control signal input to the target single-link manipulator system at time t; m0 represents the mass of the single-link manipulator; M0 is the tip load of the single-link manipulator; g is the acceleration due to gravity; l represents the length of the single-link manipulator; y k (t) represents the output signal of the target single-link manipulator system at time t; T KT represents the current time corresponding to the control signal of the target single-link manipulator system; d The expected working time of the manipulator in the target single-link manipulator system.

[0082] For example, at the initial moment of each iteration, the motor position is set to 0m, the motor speed to 0m / s, m0 and l represent the mass and length of the robot arm, respectively, with values ​​of 0.5kg and 0.5m, M0 is the tip load of the robot arm, with a value of 1kg, and g is the acceleration due to gravity, with a value of 9.8m / s². 2 .

[0083] In one embodiment, the step of optimizing the target single-link manipulator model using an ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model includes:

[0084] Initialize the target parameters of the ant colony algorithm;

[0085] Based on the target single-link manipulator model, the current angular position of the target single-link manipulator is randomly set, the fitness function for obtaining pheromones is calculated, the initial pheromones are obtained, and the state transition probability is calculated.

[0086] The initial pheromone and the state transition probability are input into the fitness function to iteratively optimize the target single-link manipulator model and obtain the updated pheromone.

[0087] Determine whether the number of iterations for the target single-link manipulator model has reached the maximum number of iterations;

[0088] If the number of iterations for the target single-link manipulator model reaches the maximum number of iterations, the current control gain of the target single-link manipulator model is taken as the optimal control gain and output.

[0089] Reference Figure 2 Specifically, the steps of the ant colony algorithm are as follows:

[0090] 1) Initialize the target parameters of the ant colony algorithm. The target parameters include the size of the ant colony, the pheromone evaporation coefficient Rho = 0.9, the transition probability P0 = 0.2, the number of iterations, etc.

[0091] 2) Search and transfer are performed based on the ant colony algorithm. By randomly placing m ants in different starting locations, the fitness function is calculated, the initial pheromone is obtained, and the state transition probability is calculated.

[0092] 3) Search for the position to be transitioned based on the state transition probability, and determine whether to transition based on the fitness function, and update the pheromone; when the determination result is to transition, update the pheromone based on the calculation result of the fitness function; when the determination result is not to transition, the updated pheromone is the same as the pheromone obtained in the previous iteration.

[0093] 4) Determine if the number of iterations has reached the maximum number of iterations. The maximum number of iterations can be set according to actual needs.

[0094] 5) If the number of iterations reaches the maximum number of iterations, output the optimal control gain and generate the control input; otherwise, continue to search, transfer and update pheromones based on the ant colony algorithm.

[0095] 6) Initialize the control inputs for the high-order iterative learning control algorithm HOILC;

[0096] 7) Use the optimal control gain to perform high-order iterative learning to obtain the output of the target single-link manipulator system.

[0097] By introducing the ant colony algorithm, the control gain in the non-uniform interval high-order iterative learning control method is further optimized, and the convergence speed is faster than that of the traditional high-order iterative learning control method.

[0098] Based on the preset sampling period of the target single-link manipulator system, the current desired output trajectory and the current control signal are determined, and the current control signal is input into the target single-link manipulator system to obtain the current output signal.

[0099] Specifically, the expression for the target single-link manipulator model in a discrete system is as follows:

[0100]

[0101]

[0102]

[0103]

[0104] C = [0 1];

[0105]

[0106] In the formula, Let be the transpose matrix of each sampled value of the target single-link manipulator model at time n+1 during the k-th iteration; Let be the transpose matrix of each sampled value of the target single-link manipulator model at time n and in the k-th iteration; This represents the control signal input to the k-th iteration of the target single-link manipulator model at time n; This represents the sampled angular position of the single-link manipulator during the k-th iteration at time n. This represents the sampled velocity value of the single-link manipulator during the k-th iteration at time n; The sampled value of the output signal of the target single-link manipulator system at time n; m0 represents the mass of the single-link manipulator; l represents the length of the single-link manipulator; M0 is the tip load of the single-link manipulator; g is the acceleration due to gravity; J represents the moment of inertia of the relative joints of the target single-link manipulator system; T S The sampling period represents the target single-link manipulator system; for example, the sampling period can be 0.05s.

[0107] Based on the actual sampling period T of the controlled continuous system S The sampled discrete sequence is:

[0108]

[0109] The corresponding discrete-time series is:

[0110]

[0111] Obtain the expected output trajectory y d (t), based on the sampling period T S Sampling is performed to obtain the expected output sequence y. d (n·T S ), Define the sequence as

[0112] For control signal u k (t) is discretized:

[0113]

[0114] Control signal u k (t) is input to the actual controlled continuous system to obtain the output signal y of the actual controlled system. k (t), based on the sampling period T S For y k (t) is sampled to obtain the output sequence y k (n·T S ), Define the sequence as follows

[0115] In one embodiment, when setting the sampling period of the target single-link manipulator system to determine the current desired output trajectory, the following steps are included:

[0116] The expression for the output signal of the target single-link manipulator system at time t is as follows:

[0117]

[0118] In the formula, y d (t) represents the desired output signal of the target single-link manipulator system at time t;

[0119] According to the preset sampling period T S For y d (t) is sampled to obtain the expected output sequence y of the target single-link manipulator system at time t. d (t·T S ), that is, the desired output trajectory of the target single-link manipulator system at time t.

[0120] Since the trajectory length of a single-link manipulator is a non-uniform interval, and since the relative degree of the single-link manipulator control system can be 1, the non-uniform interval... It can be set to [50 60]. Based on the current desired output trajectory and the current output signal, the sampling error function of the target single-link manipulator system is designed, that is, based on the sampled output sequence. Combined with the desired output trajectory sequence Define an error sequence to reduce the sampling error of the actual controlled continuous system.

[0121] In one embodiment, the sampling error function of the target single-link manipulator system is designed based on the current desired output trajectory and the current output signal using the following formula:

[0122]

[0123] E{λ k (n)}=q k (n);

[0124]

[0125]

[0126] In the formula, This represents the sampling error value of the target single-link manipulator system at time n during the k-th iteration; The target single-link manipulator system represents the expected output trajectory sample value at time n; λ represents the sampled output signal value of the target single-link manipulator system at time n; k qk(n) represents a Bernoulli random variable at time n, taking values ​​of 0 and 1; qk(n) represents a Bernoulli random variable λ. k(n) represents the expected value at time n; ek(n) represents the error value of the target single-link manipulator system in the k-th iteration at time n; T S T represents the sampling period of the target single-link manipulator system; d The expected working time of the manipulator representing the target single-link manipulator system can be set to 2.7s; T K This represents the current moment corresponding to the control signal of the target single-link manipulator system.

[0127] In one embodiment, when designing the iterative learning control law for the target single-link manipulator model based on the sampling error function and the optimal control gain, the following formula is used:

[0128]

[0129]

[0130] |W v -L v C(n+1)B(n)|≤γ v ;

[0131]

[0132]

[0133] C = [0 1];

[0134] In the formula, V represents the sampled value of the control signal input to the target single-link manipulator model at time n during the (k+1)th iteration; V represents the higher order, ranging from 1, 2, ..., N; W v and L v W represents the optimal control gain for the ant colony algorithm. v It is the optimal control gain (also called the optimal weighting coefficient of the control input) acting on the control input, L v It is the optimal control gain acting on the sampling error function (also called the optimal weighting coefficient of the sampling error function); This represents the sampled value of the control signal input to the target single-link manipulator model at time n during the (k-V+1)th iteration. T represents the sampling error value of the target single-link manipulator system at time n in the (k-V+1)th iteration; S T represents the sampling period of the target single-link manipulator system; d J represents the expected working time of the manipulator in the target single-link manipulator system; J represents the moment of inertia of the relative joints of the target single-link manipulator system.

[0135] in,

[0136]

[0137] |W v -L v C(n+1)B(n)|≤γ v ;

[0138]

[0139] The control gain W, calculated using the ant colony algorithm, represents the higher-order iterative learning control. v and L v When the convergence condition is met, the optimal control gain is obtained, meaning that the high-order iterative learning algorithm based on the ant colony algorithm calculates the optimal solution at this point. γ v Let any number satisfy the following condition:

[0140] |W v -L v C(n+1)B(n)|≤γ v ;

[0141]

[0142] Furthermore, the target single-link manipulator model is controlled to perform high-order iterative learning, and the control signal for the next iteration of the target single-link manipulator model is updated using the iterative learning control law until the sampling error function converges and the iteration stops, thus obtaining the current control signal for controlling the target single-link manipulator system.

[0143] Compared to first-order iterative learning control algorithms, high-order iterative learning control algorithms (HOILC) update the next input by learning the tracking information from the previous few iterations, resulting in better tracking performance than first-order iterative learning control algorithms.

[0144] In one embodiment, when updating the control signal for the next iteration of the target single-link manipulator model using the iterative learning control law, the following formula is employed:

[0145]

[0146] In the formula, u k+1 (t) represents the control signal of the target single-link manipulator model at time t for the (k+1)th iteration; The sampled value of the control signal for the (k+1)th iteration of the target single-link manipulator model at time t; T S T represents the sampling period of the target single-link manipulator system; d The expected working time of the manipulator in the target single-link manipulator system.

[0147] In one embodiment, when the sampling error function converges, it converges to within the tolerance error, and the expression for the error index function of the tolerance error is as follows:

[0148]

[0149] In the formula, ME k The value of the function representing the error index; y d (n) represents the expected output trajectory at time n; y k (n) represents the output signal of the target single-link manipulator system at time n; T d T represents the expected working time of the manipulator in the target single-link manipulator system. S The sampling period represents the target single-link manipulator system.

[0150] The iteration stops when the tracking error of the sampling error function converges to the range of the error index function value of the tolerance error, that is, when the sampling error function is less than the allowable range.

[0151] In this embodiment, it is assumed that the robot arm is initially stationary, so the initial velocity is 0 m / s, and the sampling period T of the controlled continuous system is set. S Secondly, obtain the desired output value of the robot arm sampling; then set the initial continuous input signals u0(t) and u1(t) of the controlled system, with both u0(t) and u1(t) set to 0, and input the signals u0(t) and u1(t) to the actual controlled system to obtain the current output signal of the actual system.

[0152] Based on the current expected output trajectory and the current output signal of the actual system, design the sampling error function sequence of the system.

[0153] Finally, based on the sampling error function sequence and the optimal control gain obtained through ant colony optimization iteration, the iterative learning control law of the model is designed.

[0154] Based on the learning control law, the control model performs high-order iterative learning, updates the control signal of the system in the next iteration, until the error index gradually converges, obtains the current control signal of the model and outputs it to control the actual robot control end, so that the system output can completely track the expected output.

[0155] The control gains obtained by iterating through the ant colony algorithm 10 times are shown in Table 1 below. The control gain of the last iteration is selected as the optimal control gain and substituted into the learning law.

[0156] Table 1

[0157]

[0158] Reference Figure 3The correspondence between the system output and the expected output trajectory in the 5th, 10th, and 30th iterations of this application is shown. It can be seen that the system's output trajectory gradually tracks the expected output trajectory as the number of iterations increases.

[0159] Reference Figure 4 The error index of this application converges within a finite number of iterations. As the number of iterations increases, the error index almost converges in the 40th iteration, meaning that the output trajectory of the robot almost matches the expected output trajectory. Using the iterative learning control algorithm of this application, the error index can converge to within the tolerance error in the 40th iteration, while traditional high-order iterative learning control laws require 100 iterations before the error index converges to within the tolerance error.

[0160] Compared with traditional high-order iterative learning control methods, which do not obtain learning gains W and L through ant colony optimization algorithms, this application reduces the number of iterations by about 60 while keeping the output error within an acceptable range. This can accelerate the convergence speed while ensuring a reduction in tracking error, reduce system runtime, and avoid unnecessary energy consumption.

[0161] In summary, a high-order iterative learning control method based on ant colony optimization for non-uniform intervals in single-link manipulators is proposed. This method constructs a target single-link manipulator model to represent the actual single-link manipulator system. The ant colony optimization algorithm is used to optimize the target single-link manipulator model, resulting in improved control gain in the iterative learning control method for non-uniform intervals. A sampling period for the target single-link manipulator system is preset to determine the current desired output trajectory and the current control signal. The current control signal is then input into the target single-link manipulator system to obtain the current output signal. Based on the current desired output trajectory and the current output signal, a sampling error function for the target single-link manipulator system is designed to correct the control signal error during the iteration process, thereby improving the control accuracy of the target single-link manipulator system. Based on sampling... Based on the error function and optimal control gain, an iterative learning control law is designed for the target single-link manipulator model. This law enables high-order iterative learning of the target single-link manipulator model, and updates the control signal for the next iteration using this iterative learning control law. This helps the target single-link manipulator model learn the optimal control signal, resulting in better system output. Compared to first-order iterative learning control algorithms, this application's high-order iterative learning control algorithm (HOILC) updates the model's next input by learning the tracking information from previous iterations. Therefore, its tracking performance is superior to that of first-order iterative learning control algorithms, and its convergence speed is faster than traditional high-order iterative learning control methods. This application achieves higher control accuracy for the single-link manipulator, exhibiting excellent performance indicators and wider applicability.

[0162] In practical applications, this application can complete the tracking task with a faster convergence speed under different trajectory lengths of single-link manipulators, effectively reducing system running time, improving production efficiency, avoiding energy loss, and greatly saving production costs.

[0163] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0164] Reference Figure 5 This application also provides a high-order iterative learning control device based on ant colony algorithm for non-consistent regions of a single-link manipulator. This device corresponds one-to-one with the high-order iterative learning control method based on ant colony algorithm for non-consistent regions of a single-link manipulator described in the above embodiments. The high-order iterative learning control device based on ant colony algorithm for non-consistent regions of a single-link manipulator includes...

[0165] The system model module is used to obtain the current angular position and current velocity of the target single-link manipulator system, and to construct the target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia and expected working time of the target single-link manipulator system.

[0166] The control gain optimization module is used to optimize the target single-link manipulator model using the ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model.

[0167] The system output module is used to preset the sampling period of the target single-link manipulator system to determine the current desired output trajectory and the current control signal, and input the current control signal into the target single-link manipulator system to obtain the current output signal;

[0168] The system error module is used to design the sampling error function of the target single-link manipulator system based on the current expected output trajectory and the current output signal.

[0169] The control law module is used to design the iterative learning control law of the target single-link manipulator model based on the sampling error function and the optimal control gain.

[0170] An iterative module is used to control the target single-link manipulator model to perform high-order iterative learning, and to update the control signal of the target single-link manipulator model for the next iteration using the iterative learning control law, until the sampling error function converges and the iteration stops, and the current control signal is obtained to control the target single-link manipulator system.

[0171] For specific limitations on a high-order iterative learning control device based on ant colony algorithm applied to the non-uniform region of a single-link manipulator, please refer to the limitations on any high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator mentioned above, which will not be repeated here.

[0172] The modules in the aforementioned high-order iterative learning control device based on ant colony algorithm for non-uniform regions of a single-link manipulator can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0173] In one embodiment, a computer device is provided, which may be a server. The computer device includes a processor, memory, a network interface, and a database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements any of the above-described high-order iterative learning control methods based on ant colony algorithms applied to the non-uniform region of a single-link manipulator.

[0174] In one embodiment, a computer-readable storage medium is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-described high-order iterative learning control methods based on ant colony algorithm applied to the non-uniform region of a single-link manipulator.

[0175] In one embodiment, a computer program product is provided, comprising a computer program that, when executed by a processor, implements any of the above-described high-order iterative learning control methods based on ant colony algorithm applied to the non-uniform region of a single-link manipulator.

[0176] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. When executed, the computer program may include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0177] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

Claims

1. A high-order iterative learning control method based on ant colony algorithm applied to non-uniform regions of a single-link manipulator, characterized in that, Includes the following steps, Obtain the current angular position and current velocity of the target single-link manipulator system, and construct a target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia, and expected working time of the target single-link manipulator system. The target single-link manipulator model is optimized using an ant colony algorithm to obtain the optimal control gain. The steps include: initializing the target parameters of the ant colony algorithm; randomly setting the current angular position of the target single-link manipulator based on the target single-link manipulator model, calculating the fitness function for acquiring pheromones to obtain initial pheromones, and calculating the state transition probability; inputting the initial pheromones and the state transition probability into the fitness function for iterative optimization of the target single-link manipulator model to obtain updated pheromones; determining whether the number of iterations of the target single-link manipulator model has reached the maximum number of iterations; if the number of iterations of the target single-link manipulator model has reached the maximum number of iterations, the current control gain of the target single-link manipulator model is taken as the optimal control gain and output. The sampling period of the target single-link manipulator system is preset to determine the current desired output trajectory and the current control signal, and the current control signal is input into the target single-link manipulator system to obtain the current output signal. Based on the current desired output trajectory and the current output signal, the sampling error function of the target single-link manipulator system is designed; the following formula is used when designing the sampling error function of the target single-link manipulator system based on the current desired output trajectory and the current output signal. In the formula, This represents the sampling error value of the target single-link manipulator system at time n during the k-th iteration; The target single-link manipulator system represents the expected output trajectory sample value at time n; The output signal sample value of the target single-link manipulator system at time n; Let n be a Bernoulli random variable at time n, taking the values ​​0 and 1; Represents a Bernoulli random variable The expected value at time n; This represents the error value of the target single-link manipulator system at time n during the k-th iteration; The sampling period represents the target single-link manipulator system; The expected working time of the manipulator representing the target single-link manipulator system; The time corresponding to the current control signal of the target single-link manipulator system; Based on the sampling error function and the optimal control gain, an iterative learning control law for the target single-link manipulator model is designed. When designing the iterative learning control law for the target single-link manipulator model based on the sampling error function and the optimal control gain, the following formula is used. In the formula, This represents the nth time step of the target single-link manipulator model. The sampled value of the control signal input in the next iteration; Represents the order of higher orders, with a range of ; This represents the optimal control gain of the ant colony algorithm applied to the control input. This represents the optimal control gain of the ant colony algorithm applied to the sampling error function; This represents the nth time step of the target single-link manipulator model. The sampled value of the control signal input in the next iteration; The target single-link manipulator system at time n is represented by the first... The sampling error value of the next iteration; The sampling period represents the target single-link manipulator system; The expected working time of the manipulator representing the target single-link manipulator system; The moment of inertia of the relative joints of the target single-link manipulator system; The target single-link manipulator model is controlled to perform high-order iterative learning, and the control signal for the next iteration of the target single-link manipulator model is updated using the iterative learning control law until the sampling error function converges and the iteration stops. The current control signal is then obtained to control the target single-link manipulator system.

2. The high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator according to claim 1, characterized in that, When updating the control signal for the next iteration of the target single-link manipulator model using the iterative learning control law, the following formula is used: In the formula, The control signal representing the (k+1)th iteration of the target single-link manipulator model at time t; The sampled value of the control signal for the (k+1)th iteration of the target single-link manipulator model at time t; The sampling period represents the target single-link manipulator system; The expected working time of the manipulator in the target single-link manipulator system.

3. The high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator according to any one of claims 1 to 2, characterized in that, The expression for the target single-link manipulator model is as follows: In the formula, For a single-link manipulator at time t The sampled value of the angle position in the next iteration; For a single-link manipulator at time t The sampled value of the velocity in the next iteration; For a single-link manipulator at time t The speed of each iteration; For a single-link manipulator at time t The angle position of the next iteration; The moment of inertia of the relative joints of the target single-link manipulator system; This represents the control signal input to the target single-link manipulator system at time t; This represents the mass of a single-link robotic arm. For the tip load of a single-link manipulator; It is the acceleration due to gravity; This represents the length of a single-link robotic arm; The output signal of the target single-link manipulator system at time t; The time corresponding to the current control signal of the target single-link manipulator system; The expected working time of the manipulator in the target single-link manipulator system.

4. The high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator according to claim 3, characterized in that, When the sampling error function converges, it converges to within the tolerance error. The expression for the error index function of the tolerance error is as follows. In the formula, The function value representing the error index function; The expected output trajectory of the target single-link manipulator system at time n; The output signal of the target single-link manipulator system at time n; The expected working time of the manipulator representing the target single-link manipulator system; The sampling period represents the target single-link manipulator system.

5. The high-order iterative learning control method based on ant colony algorithm applied to the non-uniform region of a single-link manipulator according to claim 3, characterized in that, When setting the sampling period of the target single-link manipulator system to determine the current desired output trajectory, the following steps are included: The expression for the output signal of the target single-link manipulator system at time t is as follows: In the formula, The expected output signal of the target single-link manipulator system at time t; According to the preset sampling period right Sampling is performed to obtain the desired output sequence of the target single-link manipulator system at time t. That is, the expected output trajectory of the target single-link manipulator system at time t. .

6. A high-order iterative learning control device based on ant colony algorithm for non-uniform regions of a single-link manipulator, characterized in that, include, The system model module is used to obtain the current angular position and current velocity of the target single-link manipulator system, and to construct the target single-link manipulator model by combining the mass, length, tip load, gravitational acceleration, relative joint moment of inertia and expected working time of the target single-link manipulator system. The control gain optimization module is used to optimize the target single-link manipulator model using an ant colony algorithm to obtain the optimal control gain of the target single-link manipulator model. Specifically, it initializes the target parameters of the ant colony algorithm; based on the target single-link manipulator model, it randomly sets the current angular position of the target single-link manipulator, calculates the fitness function for obtaining pheromones, obtains the initial pheromone, and calculates the state transition probability; it inputs the initial pheromone and the state transition probability into the fitness function to iteratively optimize the target single-link manipulator model to obtain the updated pheromone; it determines whether the number of iterations of the target single-link manipulator model has reached the maximum number of iterations; if the number of iterations of the target single-link manipulator model has reached the maximum number of iterations, it outputs the current control gain of the target single-link manipulator model as the optimal control gain. The system output module is used to preset the sampling period of the target single-link manipulator system to determine the current desired output trajectory and the current control signal, and input the current control signal into the target single-link manipulator system to obtain the current output signal; The system error module is used to design the sampling error function of the target single-link manipulator system based on the current desired output trajectory and the current output signal. The following formula is used when designing the sampling error function of the target single-link manipulator system based on the current desired output trajectory and the current output signal. In the formula, This represents the sampling error value of the target single-link manipulator system at time n during the k-th iteration; The target single-link manipulator system represents the expected output trajectory sample value at time n; The output signal sample value of the target single-link manipulator system at time n; Let n be a Bernoulli random variable at time n, taking the values ​​0 and 1; Represents a Bernoulli random variable The expected value at time n; This represents the error value of the target single-link manipulator system at time n during the k-th iteration; The sampling period represents the target single-link manipulator system; The expected working time of the manipulator representing the target single-link manipulator system; The time corresponding to the current control signal of the target single-link manipulator system; The control law module is used to design the iterative learning control law of the target single-link manipulator model based on the sampling error function and the optimal control gain; the following formula is used when designing the iterative learning control law of the target single-link manipulator model based on the sampling error function and the optimal control gain. In the formula, This represents the nth time step of the target single-link manipulator model. The sampled value of the control signal input in the next iteration; Represents the order of higher orders, with a range of ; This represents the optimal control gain of the ant colony algorithm applied to the control input. This represents the optimal control gain of the ant colony algorithm applied to the sampling error function; This represents the nth time step of the target single-link manipulator model. The sampled value of the control signal input in the next iteration; The target single-link manipulator system at time n is represented by the first... The sampling error value of the next iteration; The sampling period represents the target single-link manipulator system; The expected working time of the manipulator representing the target single-link manipulator system; The moment of inertia of the relative joints of the target single-link manipulator system; An iterative module is used to control the target single-link manipulator model to perform high-order iterative learning, and to update the control signal of the target single-link manipulator model for the next iteration using the iterative learning control law, until the sampling error function converges and the iteration stops, and the current control signal is obtained to control the target single-link manipulator system.

7. A computer device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method according to any one of claims 1 to 5.

Images