A nonlinear global sliding mode finite time neural network control method for gear inspection
By building a modelless control framework in the gear inspection robot system and using a radial-based neural network observer for real-time compensation, nonlinear global sliding mode surface combined with finite time convergence and adaptive approach law for system disturbance, nonlinear global sliding is designed. The mode finite time neural network controller solves the problems of low trajectory tracking and poor repeat positioning accuracy of the gear patrol robot system, achieving higher control accuracy and robustness.
Patent Information
- Application Number
- CN202310336510.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-03-31
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2043-03-31
AI Technical Summary
The gear patrol robot system has low trajectory tracking control accuracy and poor repeat positioning accuracy.
Based on the mathematical model of the N-joint robot arm of the gear patrol robot system, a model-free control framework is constructed and the tracking error equation is defined. The radial-based neural network observer is used to perform real-time observation compensation of uncertain parameters and external disturbances, combined with finite time convergence non- Design of a nonlinear global sliding mode finite time neural network controller for linear global sliding mode surface and system perturbation adaptive approach law.
It improves control accuracy and robustness, enhances adaptability in uncertain environments, and solves the problem of poor repeat positioning accuracy.
Smart Images

Figure CN116954063B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of industrial robot control systems, and in particular to a gear inspection nonlinear global sliding mode finite time neural network control method. Background Art
[0002] Inspection robots are devices with multi-dimensional autonomous movement and vision systems, such as tunnel inspection robots, power inspection robots, and security inspection robots. Among them, the gear inspection robot is a device used for high-precision gear processing technology inspection. The gear inspection robot consists of a multi-joint robotic arm and a vision system, and its own control system can achieve real-time tracking of the multi-joint robotic arm trajectory. However, the gear inspection robot has problems such as low existing control accuracy and poor repeatability due to the uncertainty of external disturbances and the uncertainty of its own model parameters. Summary of the invention
[0003] In view of the above existing problems, the present invention is proposed.
[0004] Therefore, the technical problem solved by the present invention is: how to solve the problems of low trajectory tracking control accuracy and repeated positioning accuracy of the N-joint mechanical arm of the gear inspection robot system.
[0005] To solve the above technical problems, the present invention provides the following technical solutions: based on the mathematical model of the N-joint mechanical arm of the gear inspection robot system, a model-free control framework is constructed and the tracking error equation of the N-joint mechanical arm of the gear inspection robot system is defined; a radial basis function neural network observer is used to realize real-time observation and compensation of uncertainty parameters and unknown external disturbances of the gear inspection robot system; based on the tracking error equation and nonlinear function of the N-joint mechanical arm of the gear inspection robot system, a finite-time convergent nonlinear global sliding mode surface is constructed in combination with the inertia matrix of the N-joint mechanical arm of the gear inspection robot system, the proportional term of the nonlinear function of the tracking error of the N-joint mechanical arm of the gear inspection robot system, the integral term of the nonlinear function of the tracking error of the N-joint mechanical arm of the gear inspection robot system, and the initial term of the nonlinear function of the tracking error of the N-joint mechanical arm of the gear inspection robot system; and a nonlinear global sliding mode finite-time neural network controller τ(t) of the N-joint mechanical arm of the gear inspection robot system is designed by adopting the adaptive reaching law of system disturbance.
[0006] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0007] The gear inspection robot system comprises: an image computing center, a high-power camera, and an N-joint mechanical arm.
[0008] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0009] Mathematical model of the N-joint mechanical arm of the gear inspection robot system:
[0010]
[0011] The left side of the equation is: The inertial force of the N-joint manipulator of the gear inspection robot system Centrifugal force and Coriolis force term of N-joint manipulator of gear inspection robot system The gravity term G(q(t))∈R of the N-joint manipulator of the gear inspection robot system n×1 , Gear inspection robot system N joint mechanical arm friction item External disturbance term τ d (t)∈R n×1 ; M(q(t))∈R n×n is the inertia matrix of the N-joint manipulator of the gear inspection robot system, is the centrifugal force and Coriolis force matrix of the N-joint manipulator of the gear inspection robot system; the right side of the equation is the control force term τ(t)∈R of the N-joint manipulator of the gear inspection robot system n×1 ;
[0012] Based on the mathematical model of the N-joint manipulator of the gear inspection robot system, a model-free control framework is constructed:
[0013]
[0014] Where m(q(t))∈R n×n is the non-physical parameter gain matrix of the model-free frame controller of the N-joint manipulator of the gear inspection robot system, d(t)∈R n×1 are the uncertainty parameters and unknown external disturbances of the gear inspection robot system, which are defined as follows:
[0015]
[0016] The tracking error of the N-joint manipulator of the gear inspection robot system is defined as:
[0017] e(t)=q * (t)-q(t)
[0018] Among them, q * (t) is the target trajectory of the N-joint mechanical arm of the gear inspection robot system, q(t) is the actual trajectory of the N-joint mechanical arm of the gear inspection robot system, and e(t) is the tracking error of the N-joint mechanical arm of the gear inspection robot system;
[0019] Taking the second-order differential of the tracking error of the N-joint manipulator of the gear inspection robot system, we get:
[0020]
[0021] in, is the second-order differential of e(t); Yes * The second-order differential of (t); is the second-order differential of q(t);
[0022] Define the tracking error equation of the N-joint manipulator of the gear inspection robot system:
[0023]
[0024] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0025] Define the radial basis neural network observer:
[0026]
[0027]
[0028]
[0029] w * =argmin(d(t))
[0030] h(x)=[h1(x) h2(x) … h n (x)]
[0031]
[0032] When || xo n ||≤h n :
[0033]
[0034] When h n <||xo n ||≤2h n :
[0035]
[0036] When 2h n <||xo n ||:
[0037] h n (x) = 0
[0038] in, is the estimated value of the uncertainty parameters and unknown external disturbances of the gear inspection robot system, w * is the optimal weight of the radial basis neural network observer, h(x) is the hidden layer function of the radial basis neural network observer, x is the input of the radial basis neural network observer, o n is the center of the radial basis neural network observer; h n is the bandwidth of the radial basis neural network observer.
[0039] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0040] Define the finite time convergent nonlinear global sliding surface:
[0041]
[0042] s(t)=[s1(t) s2(t) … s n (t)] T
[0043] k gnp =diag[k 1gnp k 2gnp … k ngnp ] T
[0044] k gni =diag[k 1gni k 2gni … k ngni ] T
[0045]
[0046] k=diag[k1 k2 … k n ] T
[0047] fal[e(t),γ,η]
[0048] =[fal[e1(t),γ,η] fal[e2(t),γ,η] … fal[e n (t),γ,η]] T
[0049]
[0050] fal[e(0),γ,η]
[0051] =[fal[e1(0),γ,η] fal[e2(0),γ,η] … fal[e n (0),γ,η]] T
[0052]
[0053]
[0054]
[0055] Among them, k gnp fal[e(t),γ,η] is the proportional term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, k gnp is the proportionality coefficient, is the integral term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, k gni is the integral coefficient, k is the feedback gain, k gnp fal[e(0),γ,η] is the initial term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system;
[0056] Take the first-order differential of the finite-time convergent nonlinear global sliding surface:
[0057]
[0058]
[0059]
[0060] The limited time periods defined are:
[0061] When |e(t)|>η:
[0062]
[0063] When |e(t)|<η:
[0064]
[0065] Among them, t s is a finite convergence time, and C is an arbitrary constant.
[0066] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0067] Define the system disturbance adaptive reaching law:
[0068]
[0069]
[0070] Among them, δ is the adaptive reaching law adjustment gain of the system disturbance, and its adaptive rate is β1, β2, …, β n is the tuning gain,
[0071] As a preferred solution of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, wherein:
[0072] Based on the radial basis neural network observer, the finite time convergent nonlinear global sliding mode surface, and the system disturbance adaptive reaching law, the nonlinear global sliding mode finite time neural network controller τ(t) of the N-joint manipulator of the gear inspection robot system is designed by the model-free control framework:
[0073] When |e(t)|>η:
[0074]
[0075] When |e(t)|<η:
[0076]
[0077] The beneficial effects of the present invention are as follows: (1) The method of the present invention is based on the mathematical model of the N-joint mechanical arm of the gear inspection robot system, constructs a model-free control framework and defines the tracking error equation of the N-joint mechanical arm of the gear inspection robot system; on this basis, a radial basis neural network observer is proposed to realize compensation for system uncertainty; (2) The inertia matrix of the N-joint mechanical arm of the gear inspection robot system, the proportional term of the nonlinear function of the tracking error of the N-joint mechanical arm of the gear inspection robot system, the integral term of the nonlinear function, and the initial term of the nonlinear function are used to construct a finite-time convergent nonlinear global sliding surface to enhance the control robustness; (3) The controller is designed using the system disturbance adaptive convergence law to realize the adaptability of the approaching stage. BRIEF DESCRIPTION OF THE DRAWINGS
[0078] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.
[0079] in:
[0080] Figure 1A control principle diagram of a nonlinear global sliding mode finite time neural network control method for gear inspection according to an embodiment of the present invention;
[0081] Figure 2 A schematic diagram of the target tracking trajectory (aim1) of a dual-joint mechanical arm of a gear inspection robot system and a trajectory tracking method of a nonlinear global sliding mode finite-time neural network control method for gear inspection (NGSMFTNNC) according to an embodiment of the present invention;
[0082] Figure 3 A schematic diagram of the target tracking trajectory (aim2) of a dual-joint mechanical arm of a gear inspection robot system and a trajectory tracking method of a nonlinear global sliding mode finite-time neural network control method for gear inspection (NGSMFTNNC) according to an embodiment of the present invention;
[0083] Figure 4 A schematic diagram of the target tracking trajectory (aim1) of a dual-joint mechanical arm of a gear inspection robot system and a trajectory tracking method of a model-based neural network control method (NNFC) for a gear inspection nonlinear global sliding mode finite-time neural network control method according to an embodiment of the present invention;
[0084] Figure 5 A schematic diagram of the target tracking trajectory (aim2) of a dual-joint mechanical arm of a gear inspection robot system and a trajectory tracking method of a model-based neural network control method (NNFC) for a gear inspection nonlinear global sliding mode finite-time neural network control method according to an embodiment of the present invention; DETAILED DESCRIPTION
[0085] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings.
[0086] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0087] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.
[0088] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.
[0089] Reference Figure 1 , is an embodiment of the present invention, which provides a nonlinear global sliding mode finite time neural network control method for gear inspection. The method of the present invention organically integrates the mathematical model of the N-joint mechanical arm of the gear inspection robot system, the radial basis neural network observer, the finite time convergence nonlinear global sliding mode surface, and the system disturbance adaptive reaching law to design the nonlinear global sliding mode finite time neural network controller τ(t) of the N-joint mechanical arm of the gear inspection robot system. Figure 1 , is a control block diagram of the gear inspection nonlinear global sliding mode finite time neural network control method of the present invention, specifically including:
[0090] S1: Based on the mathematical model of the N-joint robotic arm of the gear inspection robot system, a model-free control framework is constructed and the tracking error equation of the N-joint robotic arm of the gear inspection robot system is defined.
[0091] The gear inspection robot system comprises: an image computing center, a high-power camera, and an N-joint mechanical arm.
[0092] Mathematical model of the N-joint mechanical arm of the gear inspection robot system:
[0093]
[0094] The left side of the equation is: The inertial force of the N-joint manipulator of the gear inspection robot system Centrifugal force and Coriolis force term of N-joint manipulator of gear inspection robot system The gravity term G(q(t))∈R of the N-joint manipulator of the gear inspection robot system n×1 , Gear inspection robot system N joint mechanical arm friction item External disturbance term τ d (t)∈R n×1 ; M(q(t))∈R n×n is the inertia matrix of the N-joint manipulator of the gear inspection robot system, is the centrifugal force and Coriolis force matrix of the N-joint manipulator of the gear inspection robot system; the right side of the equation is the control force term τ(t)∈R of the N-joint manipulator of the gear inspection robot system n×1 .
[0095] Based on the mathematical model of the N-joint manipulator of the gear inspection robot system, a model-free control framework is constructed:
[0096]
[0097] Where m(q(t))∈R n×n is the non-physical parameter gain matrix of the model-free frame controller of the N-joint manipulator of the gear inspection robot system, d(t)∈R n×1 are the uncertainty parameters and unknown external disturbances of the gear inspection robot system, which are defined as follows:
[0098]
[0099] The tracking error of the N-joint manipulator of the gear inspection robot system is defined as:
[0100] e(t)=q * (t)-q(t
[0101] Among them, q * (t) is the target trajectory of the N-joint robotic arm of the gear inspection robot system, q(t) is the actual trajectory of the N-joint robotic arm of the gear inspection robot system, and e(t) is the tracking error of the N-joint robotic arm of the gear inspection robot system.
[0102] Taking the second-order differential of the tracking error of the N-joint manipulator of the gear inspection robot system, we get:
[0103]
[0104] in, is the second-order differential of e(t); Yes * The second-order differential of (t); is the second-order differential of q(t).
[0105] Define the tracking error equation of the N-joint manipulator of the gear inspection robot system:
[0106]
[0107] S2: Use the radial basis function neural network observer to realize real-time observation and compensation of the uncertainty parameters and unknown external disturbances of the gear inspection robot system.
[0108] Define the radial basis neural network observer:
[0109]
[0110]
[0111]
[0112] w* =argmin(d(t))
[0113] h(x)=[h1(x) h2(x) … h n (x)]
[0114]
[0115] When || xo n ||≤h n :
[0116]
[0117] When h n <||xo n ||≤2h n :
[0118]
[0119] When 2h n <||xo n ||:
[0120] h n (x) = 0
[0121] in, is the estimated value of the uncertainty parameters and unknown external disturbances of the gear inspection robot system, w * is the optimal weight of the radial basis neural network observer, h(x) is the hidden layer function of the radial basis neural network observer, x is the input of the radial basis neural network observer, o n is the center of the radial basis neural network observer; h n is the bandwidth of the radial basis neural network observer.
[0122] S3: Based on the tracking error equation and nonlinear function of the N-joint robotic arm of the gear inspection robot system, combined with the inertia matrix of the N-joint robotic arm of the gear inspection robot system, the proportional term of the nonlinear function of the tracking error of the N-joint robotic arm of the gear inspection robot system, the integral term of the nonlinear function of the tracking error of the N-joint robotic arm of the gear inspection robot system, and the initial term of the nonlinear function of the tracking error of the N-joint robotic arm of the gear inspection robot system, a finite-time convergent nonlinear global sliding surface is constructed.
[0123] Define the finite time convergent nonlinear global sliding surface:
[0124]
[0125] s(t)=[s1(t) s2(t) … s n (t)]T
[0126] k gnp =diag[k 1gnp k 2gnp … k ngnp ] T
[0127] k gni =diag[k 1gni k 2gni …k ngni ] T
[0128]
[0129] k=diag[k1 k2 … k n ] T
[0130] [fal[e(t),γ,η]
[0131] =[fal[e1(t),γ,η] fal[e2(t),γ,η] … fal[e n (t),γ,η]] T
[0132]
[0133]
[0134] fal[e(0),γ,η]
[0135] =[fal[e1(0),γ,η] fal[e2(0),γ,η] … fal[e n (0),γ,η]] T
[0136]
[0137]
[0138]
[0139] Among them, k gnp fal[e(t),γ,η] is the proportional term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, k gnp is the proportionality coefficient, is the integral term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, k gni is the integral coefficient, k is the feedback gain, Kgnp fal[e(0),γ,η] is the initial term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system.
[0140] Take the first-order differential of the finite-time convergent nonlinear global sliding surface:
[0141]
[0142]
[0143]
[0144]
[0145] The limited time periods defined are:
[0146] When |e(t)|>η:
[0147]
[0148] When |e(t)|<η:
[0149]
[0150] Among them, t s is a finite convergence time, and C is an arbitrary constant.
[0151] S4: Using the adaptive reaching law of system disturbance, a nonlinear global sliding mode finite-time neural network controller τ(t) for the N-joint manipulator of the gear inspection robot system is designed, and its stability is verified.
[0152] Define the system disturbance adaptive reaching law:
[0153]
[0154]
[0155] Among them, δ is the adaptive reaching law adjustment gain of the system disturbance, and its adaptive rate is β1, β2, …, β n is the tuning gain,
[0156] Based on the radial basis neural network observer, the finite time convergent nonlinear global sliding mode surface, and the system disturbance adaptive reaching law, the nonlinear global sliding mode finite time neural network controller τ(t) of the N-joint manipulator of the gear inspection robot system is designed by the model-free control framework:
[0157] When |e(t)|>η:
[0158]
[0159] When |e(t)|<η:
[0160]
[0161] The Lyapunov function is:
[0162]
[0163]
[0164] Among them, β n >0,
[0165] Reference Figure 1 , is the control principle diagram of the nonlinear global sliding mode finite-time neural network control method for gear inspection, and is a further explanation of the method of the present invention. The main control diagram process is as follows: First, based on the mathematical model of the N-joint robotic arm of the gear inspection robot system, a model-free control framework is constructed and the tracking error equation of the N-joint robotic arm of the gear inspection robot system is defined; secondly, based on the tracking error equation of the N-joint robotic arm of the gear inspection robot system, a radial basis neural network observer and a finite-time convergent nonlinear global sliding mode surface are constructed; finally, the controller τ(t) is designed based on the system disturbance adaptive reaching law.
[0166] Preferably, it should be noted that, compared with the prior art, the present invention discloses a nonlinear global sliding mode finite time neural network control method for gear inspection, which aims to track the target trajectory of the N-joint manipulator of the gear inspection robot system by using a finite time convergent nonlinear global sliding mode, and use a radial basis neural network observer to predict the system disturbance in real time, and then achieve adaptive dynamic response through the adaptive convergence law of the system disturbance. Among them, based on the model-free control framework, the combination of observer compensation and sliding mode control is adopted to solve the problems of low control accuracy and poor repeated positioning accuracy.
[0167] Reference Figure 2 to Figure 5 , provides a test verification of a nonlinear global sliding mode finite time neural network control method for gear inspection, including:
[0168] In order to verify and illustrate the technical effect adopted in the present method, a model-based neural network control method (NNFC) is selected in this embodiment to conduct comparative tests with the method of the present invention, and the test results are compared by means of scientific demonstration to verify the real effect of the present method.
[0169] The gear inspection robot has uncertain external disturbances and uncertainties in its own model parameters, resulting in low control accuracy and poor repeated positioning accuracy. In order to verify that the method of the present invention has uncertainty compensation, global robustness, and adaptability in the approaching stage compared with the traditional method, a nonlinear global sliding mode finite-time neural network control method (NGSMFTNNC) will be used in this embodiment to track the target trajectory (aim1 and aim2) of the dual-joint robotic arm of the gear inspection robot system, and the output trajectory and tracking error of the N-joint robotic arm of the gear inspection robot system will be measured and compared in real time with the model-based neural network control method (NNFC).
[0170] Test environment: Reference Figure 1 , the N-joint robot arm of the gear inspection robot system is run on the simulation platform to simulate the target trajectory (aim1 and aim2) of the dual-joint robot arm of the gear inspection robot system, and the nonlinear global sliding mode finite time neural network control method (NGSMFTNNC) and the model-based neural network control method (NNFC) are used to test and obtain the test result data. All tests will be carried out by turning on the automated test equipment and using MATLAB software programming to implement the simulation test of the comparison method, and the simulation data will be obtained according to the experimental results; each method will test 3 groups of data, and each group of data will be sampled for 10s, and the input trajectory and tracking error of each group of data will be calculated, and the error will be compared with the expected target trajectory of the simulation input.
[0171] Reference Figure 2 to Figure 5 , which is a trajectory tracking diagram comparing the tracking target trajectory (aim1 and aim2) of the dual-joint robotic arm of the gear inspection robot system, the nonlinear global sliding mode finite time neural network control method (NGSMFTNNC), and the model-based neural network control method (NNFC) of the present invention.
[0172] Parameters of the dual-joint robotic arm: mass of link 1 m1 = 1 kg, length of link 1 l1 = 1 m, distance from center of mass to joint 1 l c1 =1 / 2m, moment of inertia of connecting rod 1 I1 = 1 / 12kg·m, mass of connecting rod 2 m e =3kg, the distance from link 2 to joint 2 is l ce =1m, moment of inertia of connecting rod 2 l e =2 / 5kg·m, the angle between the center of mass and joint 2 is δ e =0, friction coefficient e1=-7 / 12, acceleration due to gravity e2=9.81.
[0173] Reference Figure 2 to Figure 5 , m(q(t))=[2 3],k 1gnp =5, k 2gnp =0.02, k1gni =0.001, k 2gni =10, k1=1, k2=2, eta=3, γ=0.5, β1=0.05, β2=0.1, in Figure 2 and Figure 3 It is not difficult to find that the nonlinear global sliding mode finite-time neural network control method (NGSMFTNNC) can effectively track the target trajectory (aim1 and aim2), but the model-based neural network control method (NNFC) cannot track and has a large tracking error.
[0174] In summary, the gear inspection nonlinear global sliding mode finite-time neural network control method proposed in the present invention is superior to other methods in terms of steady-state error, thanks to: finite-time convergence of nonlinear global sliding mode surface and adaptive convergence law, which solves the problems of low control accuracy and poor repeated positioning accuracy. In addition, the model-free framework solves the uncertainty problem of the model, and the neural network performs compensation.
[0175] It should be appreciated that embodiments of the present invention may be implemented or enforced by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory. The method may be implemented in a computer program using standard programming techniques - including a non-transitory computer-readable storage medium configured with a computer program, wherein the storage medium so configured causes the computer to operate in a specific and predefined manner - according to the methods and drawings described in the specific embodiments. Each program may be implemented in a high-level procedural or object-oriented programming language to communicate with a computer system. However, if desired, the program may be implemented in an assembly or machine language. In any case, the language may be a compiled or interpreted language. In addition, the program may be run on a programmed ASIC for this purpose.
[0176] In addition, the operations of the processes described in the present invention may be performed in any suitable order unless otherwise indicated by the present invention or otherwise clearly contradicted by the context. The processes described in the present invention (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed collectively on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions that may be executed by one or more processors.
[0177] Further, the method can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, a RAM, a ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the process described herein. In addition, the machine-readable code, or part thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the present invention includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques described in the present invention, the present invention also includes the computer itself. The computer program can be applied to input data to perform the functions described in the present invention, thereby converting the input data to generate output data stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents a physical and tangible object, including a specific visual depiction of the physical and tangible object produced on a display.
[0178] The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiment. Any equivalent modifications or changes made by ordinary technicians in this field based on the contents disclosed by the present invention should be included in the protection scope recorded in the claims.
Claims
1. A nonlinear global sliding mode finite time neural network control method for gear inspection, characterized by: The steps include: S1, based on the mathematical model of the N-joint manipulator of the gear inspection robot system, a model-free control framework is constructed and the tracking error equation of the N-joint manipulator of the gear inspection robot system is defined; S2, using radial basis neural network observer, to achieve real-time observation and compensation of uncertainty parameters and unknown external disturbances of the gear inspection robot system; S3, based on the tracking error equation and nonlinear function of the N-joint manipulator of the gear inspection robot system, combined with the inertia matrix of the N-joint manipulator of the gear inspection robot system, the proportional term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, the integral term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, and the initial term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, a finite-time convergent nonlinear global sliding surface is constructed; In step S3, the finite time convergent nonlinear global sliding mode surface is defined as: s(t)=[s1(t) s2(t) … s n (t)] T k gnp =diag[k 1gnp k 2gnp … k ngnp ] T k gni =diag[k 1gni k 2gni … k ngni ] T k=diag[k1 k2 … k n ] T Among them, k gnp fal[e(t),γ,η] is the proportional term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, k gnp is the proportionality coefficient, is the integral term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, yes Variable factors, γ and η are parameter gains, k gni is the integral coefficient, k is the feedback gain, k gnp fal[e(0),γ,η] is the initial term of the nonlinear function of the tracking error of the N-joint manipulator of the gear inspection robot system, e(t) is the tracking error of the N-joint manipulator of the gear inspection robot system, M(q) is M(q(t))∈R n×n It is the inertia matrix of the N-joint manipulator of the gear inspection robot system; Take the first-order differential of the finite-time convergent nonlinear global sliding surface: The limited time mentioned in the definition is: When |e(t)|>η: When |e(t)|<η: Among them, t s is the finite convergence time, C is an arbitrary constant; S4, using the adaptive reaching law of system disturbance, design the nonlinear global sliding mode finite time neural network controller τ(t) of the N-joint manipulator of the gear inspection robot system; In step S4, the system disturbance adaptive reaching law is defined as: Among them, δ is the adaptive reaching law adjustment gain of the system disturbance, and its adaptive rate is β1, β2, …, β n is the tuning gain, In step S4, based on the radial basis neural network observer, the finite time convergent nonlinear global sliding mode surface, and the system disturbance adaptive reaching law, a nonlinear global sliding mode finite time neural network controller τ(t) of the N-joint manipulator of the gear inspection robot system is designed by a model-free control framework: When |e(t)|>η: When |e(t)|<η: Where m(q(t))∈R n×n The N-joint robot arm model-free frame controller of the gear inspection robot system has no physical meaning parameter adjustment gain matrix. Yes * The second-order differential of (t), q * (t) is the target trajectory of the N-joint manipulator of the gear inspection robot system, It is the estimated value of the uncertainty parameters and unknown external disturbances of the gear inspection robot system.
2. A gear inspection nonlinear global sliding mode finite time neural network control method according to claim 1, characterized in that: In step S1, the gear inspection robot system includes: an image computing center, a high-magnification camera, and an N-joint robotic arm.
3. The gear inspection nonlinear global sliding mode finite time neural network control method according to claim 2 is characterized by: In step S1, the mathematical model of the N-joint mechanical arm of the gear inspection robot system is: The left side of the equation is: The inertial force of the N-joint manipulator of the gear inspection robot system Centrifugal force and Coriolis force term of N-joint manipulator of gear inspection robot system The gravity term G(q(t))∈R of the N-joint manipulator of the gear inspection robot system n×1 , Gear inspection robot system N joint mechanical arm friction item External disturbance term τ d (t)∈R n×1 ; M(q(t))∈R n×n is the inertia matrix of the N-joint manipulator of the gear inspection robot system, is the centrifugal force and Coriolis force matrix of the N-joint manipulator of the gear inspection robot system; the right side of the equation is the control force term τ(t)∈R of the N-joint manipulator of the gear inspection robot system n×1 ; Based on the mathematical model of the N-joint manipulator of the gear inspection robot system, a model-free control framework is constructed: Where m(q(t))∈R n×n is the non-physical parameter gain matrix of the model-free frame controller of the N-joint manipulator of the gear inspection robot system, d(t)∈R n×1 are the uncertainty parameters and unknown external disturbances of the gear inspection robot system, which are defined as follows: The tracking error of the N-joint manipulator of the gear inspection robot system is defined as: e(t)=q * (t)-q(t) Among them, q * (t) is the target trajectory of the N-joint mechanical arm of the gear inspection robot system, q(t) is the actual trajectory of the N-joint mechanical arm of the gear inspection robot system, and e(t) is the tracking error of the N-joint mechanical arm of the gear inspection robot system; Taking the second-order differential of the tracking error of the N-joint manipulator of the gear inspection robot system, we get: in, is the second-order differential of e(t); Yes * The second-order differential of (t); is the second-order differential of q(t); Define the tracking error equation of the N-joint manipulator of the gear inspection robot system:
4. The gear inspection nonlinear global sliding mode finite time neural network control method according to claim 3 is characterized by: In step S2, the radial basis neural network observer is defined as: w * =argmin(d(t)) h(x)=[h1(x) h2(x) … h n (x)] When || xo n ||≤h n : When h n <||xo n ||≤2h n : When 2h n <||xo n ||: h n (x)=0 in, is the estimated value of the uncertainty parameters and unknown external disturbances of the gear inspection robot system, w * is the optimal weight of the radial basis neural network observer, h(x) is the hidden layer function of the radial basis neural network observer, x is the input of the radial basis neural network observer, o n is the center of the radial basis neural network observer; h n is the bandwidth of the radial basis neural network observer.
