Self-adaptive anti-noise motion planning method for mobile manipulator based on fuzzy rule

Abstract

The invention relates to the technical field of mobile mechanical arms, in particular to a mobile mechanical arm self-adaptive anti-noise motion planning method based on a fuzzy rule. Comprising the following steps: firstly, designing a motion planning scheme according to a kinematics model of the mobile mechanical arm; the motion planning scheme of the movable mechanical arm is converted into a quadratic programming problem; secondly, designing an anti-noise zero neural network model to solve a quadratic programming problem in real time; further, a fuzzy inference system is introduced to carry out optimization and adaptive adjustment on key parameters in the anti-noise zero neural network model; and finally, the result of the fuzzy inference system is transmitted to a lower computer in real time, and the mobile mechanical arm is driven to complete expected trajectory tracking and task execution. The motion planning method has good timeliness and robustness, interference of noise on motion planning can be effectively restrained, and it can still be guaranteed that the mechanical arm smoothly completes tasks in the complex noise environment.

Description

Technical Field

[0001] This invention relates to the field of mobile robotic arm technology, and in particular to an adaptive noise-resistant motion planning method for mobile robotic arms based on fuzzy rules. Background Technology

[0002] Mobile robotic arms, as an important branch of robotics technology, combine the flexible mobility of mobile platforms with the precise manipulation capabilities of multi-joint robotic arms, and have demonstrated significant application value in complex and dynamic scenarios such as disaster relief and intelligent healthcare. Existing mobile robotic arms typically consist of a mobile platform and a multi-joint robotic arm. Their working principle is as follows: a motion planning scheme is designed based on the kinematic model of the mobile robotic arm; this scheme is then converted into control commands and transmitted to a lower-level computer; the lower-level computer uses a drive algorithm to control the rotation of the wheels and the movement of the joints, ultimately achieving trajectory tracking and task execution by the end effector.

[0003] However, multi-source noise in actual working environments, such as motor vibration noise, environmental sensor noise, and external interference noise, can significantly affect the motion planning of mobile robotic arms. Existing control methods have obvious shortcomings in noise handling: on the one hand, they lack effective noise suppression mechanisms, making it difficult to counteract the interference of noise on the control system and easily leading to positional deviations in the end effector; on the other hand, their adaptive noise compensation capability is weak, and in dynamically changing noise environments, they cannot adjust control parameters in real time to adapt to environmental changes, thereby reducing the control accuracy of the mobile robotic arm and, in severe cases, even causing the preset task to fail to be completed. Summary of the Invention

[0004] In view of this, the purpose of this invention is to propose an adaptive noise-resistant motion planning method for mobile robotic arms based on fuzzy rules, so as to solve the problems of insufficient suppression and adaptive compensation under multi-source noise interference and low control accuracy in the existing motion planning of mobile robotic arms.

[0005] To achieve the above objectives, the technical solution adopted by this invention is: an adaptive noise-resistant motion planning method for a mobile robotic arm based on fuzzy rules, comprising the following steps:

[0006] S1. Design a motion planning scheme based on the kinematic model of the mobile robotic arm;

[0007] S2. Transform the motion planning scheme of the mobile robotic arm into a quadratic programming problem;

[0008] S3. Design a noise-resistant nullable neural network model to solve quadratic programming problems in real time;

[0009] S4. Introduce a fuzzy inference system to optimize and adaptively adjust key parameters in the anti-noise nullification neural network model;

[0010] S5. Transmit the results of the fuzzy inference system to the lower-level machine in real time to drive the mobile robotic arm to complete the desired trajectory tracking and task execution.

[0011] Preferably, the kinematic model of the mobile robotic arm is as follows: ,in, Mobile platforms The displacement vector of each wheel. Robotic arms The angular displacement vector of each joint. The position of the end effector. It is a non-linear mapping function.

[0012] Preferably, the motion planning scheme is as follows: ,in, For the mobile robotic arm to be in constant motion The generalized velocity vector; for The first derivative; It is the Jacobian pseudo-inverse matrix; For the end effector at time The velocity vector.

[0013] Preferably, the formula for the quadratic programming problem is:

[0014]

[0015] in, Let the objective function be the quadratic programming problem. These are constraints for a quadratic programming problem. It is a positive definite symmetric matrix, with weights allocated to joint angular velocities and wheel speeds; These are the coefficient vectors, and the weights used to help minimize the optimization error.

[0016] Preferably, the noise-resistant nullification neural network model is:

[0017]

[0018] in, , For Jacobian matrices, This is the transpose of the Jacobian matrix. For Lagrange multipliers, , , They are respectively , , The first derivative; For error feedback gain parameters, The gain parameter is used to compensate for other noise. For pose error, and ; and For noise compensation, Used to compensate for harmonic noise. Used to compensate for other noises pose error The integral term; Unknown noise; , must meet And it is a positive integer; The frequency of harmonic noise; , for The first derivative.

[0019] Preferably, in the fuzzy inference system, the fuzzy inference process is as follows: the fuzzy inference system first performs... Input , After fuzzification, fuzzy rules are applied for fuzzy reasoning to obtain a fuzzy conclusion. Then, the fuzzification is performed again to obtain... Output .

[0020] The beneficial effects of this invention are as follows: Compared with the prior art, this invention uses fuzzy rules to reason about motion-related errors and changes, combines the Jacobian pseudo-inverse to reverse-calculate the end effector's expected speed into wheel speed and joint angular velocity, and uses the Lagrange multiplier method to transform the quadratic programming problem into an equation relationship to solve, effectively suppressing multi-source noise interference, realizing adaptive adjustment of control parameters, and improving the trajectory tracking accuracy and motion planning stability of the mobile robotic arm's end effector. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the process of the present invention;

[0023] Figure 2 This is a schematic diagram of the fuzzy reasoning process of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains.

[0025] like Figure 1 As shown, an adaptive noise-resistant motion planning method for a mobile robotic arm based on fuzzy rules includes:

[0026] S1. Design a motion planning scheme based on the kinematic model of the mobile robotic arm.

[0027] Assume the mobile robotic arm has One joint, One wheel, and the mobile platform is made of Driven by a single wheel. First, the joint coordinate system is determined, and DH parameters are obtained to represent the relationship between adjacent links. The robotic arm is then connected to the mobile platform, resulting in the kinematic model of the mobile robotic arm:

[0028]

[0029] in, Mobile platforms The displacement vector of each wheel. Robotic arms The angular displacement vector of each joint. The position of the end effector. For nonlinear mapping functions; Regarding time Differentiation yields the kinematic model of the velocity layer:

[0030]

[0031] in, For the Jacobian matrix of the mobile robotic arm, The generalized velocity vector of the mobile robotic arm consists of two parts: the mobile platform. The velocity vector of each wheel and robotic arm Angular velocity vector of each joint , for The first derivative, This is the velocity vector of the end effector.

[0032] Subsequently, based on the kinematic model of the velocity layer, the desired end-effector velocity is inversely calculated into joint velocity and wheel velocity using the Jacobian pseudo-inverse. The motion planning scheme for the mobile robotic arm is designed as follows:

[0033]

[0034] in, For the mobile robotic arm to be in constant motion The generalized velocity vector; It is the Jacobian pseudo-inverse matrix; For the end effector at time The velocity vector.

[0035] S2. Transform the motion planning scheme of the mobile robotic arm into a quadratic programming problem.

[0036] To optimize the motion performance of the mobile robot, the motion planning scheme of S1 is transformed into a quadratic programming problem, with the following formula:

[0037]

[0038] in, Let the objective function be the quadratic programming problem. These are constraints for a quadratic programming problem. It is a positive definite symmetric matrix, with weights allocated to joint angular velocities and wheel speeds; These are the coefficient vectors, and the weights used to help minimize the optimization error.

[0039] To solve the above quadratic programming problem, the Lagrange multiplier method is used. This involves introducing Lagrange multipliers. We construct the Lagrange function to further transform the quadratic programming problem into an equation relationship, as shown below:

[0040]

[0041] in, , This is the transpose of the Jacobian matrix; It is a Lagrange multiplier.

[0042] S3. Design a noise-resistant nullable neural network model to solve quadratic programming problems in real time.

[0043] Various unknown noises in the environment can affect the control precision of a mobile robotic arm. These unknown noises can be addressed by... Decomposed into harmonic noise Other noises, among which Let A, B, C, and D represent the amplitude, frequency, and phase angle of the harmonic noise, respectively. , must meet And the integer is positive. Simultaneously, a noise compensation term is designed. and ,in, Used to compensate for harmonic noise, while Used to compensate for other noises pose error The integral term. Let , and The dynamic relationship is as follows:

[0044]

[0045] Define pose error as A noise-resistant nullable neural network is designed to solve the quadratic programming problem. The noise-resistant nullable neural network model is as follows:

[0046]

[0047] in, For error feedback gain parameters, This is the gain parameter for compensating for other noise.

[0048] S4. Introduce a fuzzy inference system to optimize and adaptively adjust key parameters in the noise-reducing neural network model.

[0049] The introduced fuzzy inference system is shown in the table below.

[0050]

[0051] This fuzzy inference system has There are 1 input, which are the pose error and 2 inputs. and the integral term of pose error , The outputs are respectively from the noise-resistant nullification neural network model. Key gain parameters The fuzzy rules are defined as follows: NB (negative large), NM (negative medium), Z (zero), PM (positive medium), PB (positive large), NL (negative large), PL (positive large). (See figure) As shown, the fuzzy inference process is as follows: The fuzzy inference system first performs... Input , After fuzzification, the aforementioned fuzzy rules are applied for fuzzy reasoning to obtain a fuzzy conclusion. Then, the fuzzification is performed again to obtain the final result. Output This fuzzy inference system enables the noise-resistant nullable neural network model to adaptively adjust parameters in real time according to the dynamic environment, thereby minimizing errors and achieving better noise resistance.

[0052] S5. Transmit the fuzzy inference system results to the lower-level computer in real time to drive the mobile robotic arm to complete the desired trajectory tracking and task execution.

[0053] The joint and wheel speeds of the mobile robotic arm are calculated using a fuzzy inference system, and the results are transmitted to the lower-level controller. This controller is equipped with a high-efficiency drive algorithm and multiple joint and wheel drive modules for the mobile robotic arm. The drive algorithm converts the calculation results into specific drive commands, thereby controlling the joints and wheels of the mobile robotic arm and ensuring that it can stably perform its intended tasks even in various noise interference environments.

[0054] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.

[0055] This invention aims to cover all such substitutions, modifications, and variations that fall within the scope of protection. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for adaptive noise-resistant motion planning of a mobile robotic arm based on fuzzy rules, characterized in that, Includes the following steps: S1. Design a motion planning scheme based on the kinematic model of the mobile robotic arm; S2. Transform the motion planning scheme of the mobile robotic arm into a quadratic programming problem; S3. Design a noise-resistant nullable neural network model to solve quadratic programming problems in real time; S4. Introduce a fuzzy inference system to optimize and adaptively adjust key parameters in the anti-noise nullification neural network model; S5. Transmit the results of the fuzzy inference system to the lower-level machine in real time to drive the mobile robotic arm to complete the desired trajectory tracking and task execution.

2. The method according to claim 1, characterized in that, The kinematic model of the mobile robotic arm is as follows: ,in, Mobile platforms The displacement vector of each wheel. Robotic arms The angular displacement vector of each joint. The position of the end effector. It is a non-linear mapping function.

3. The method according to claim 1, characterized in that, The motion planning scheme is as follows: ,in, For the mobile robotic arm to be in constant motion The generalized velocity vector; for The first derivative; It is the Jacobian pseudo-inverse matrix; For the end effector at time The velocity vector.

4. The method according to claim 1, characterized in that, The formula for the quadratic programming problem is: in, Let the objective function be the quadratic programming problem. These are constraints for a quadratic programming problem. It is a positive definite symmetric matrix, with weights allocated to joint angular velocities and wheel speeds; These are the coefficient vectors, and the weights used to help minimize the optimization error.

5. The method according to claim 1, characterized in that, The noise-resistant nullification neural network model is as follows: in, , For Jacobian matrices, This is the transpose of the Jacobian matrix. For Lagrange multipliers, , , They are respectively , , The first derivative; For error feedback gain parameters, The gain parameter is used to compensate for other noise. For pose error, and ; and For noise compensation, Used to compensate for harmonic noise. Used to compensate for other noise and for The integral term; Unknown noise; , must meet And it is a positive integer; The frequency of harmonic noise; , for The first derivative.

6. The method according to claim 1, characterized in that, The fuzzy reasoning system, wherein the fuzzy reasoning process is as follows: the fuzzy reasoning system first performs... Input , After fuzzification, fuzzy rules are applied for fuzzy reasoning to obtain a fuzzy conclusion. Then, the fuzzification is performed again to obtain... Output .

Images