In this work, a new method based on the nonlinear output frequency response functions (NOFRFs) and deep convolution neural network (CNN) is proposed for RV reducer fault diagnosis. In order to solve the problem of low accuracy caused by output signal that cannot describe the nonlinear characteristics of system, the nonlinear spectrum based on output frequency response function is adopted. To improve the performance of error cost function model in back propagation (BP) process of CNN, a new comprehensive error cost function model is designed to guide the network parameters' optimization in direction of the feature classification. Besides, an adaptive network structure optimization algorithm is proposed to avoid the blindness of network structure selection by experience. Specifically, 4-order NOFRF spectrums of system are first obtained by adaptive identification algorithm. Next, NOFRF spectrums are transformed into 2-D images, and CNN extracts the fault features hidden in spectrum images and performs fault classification. The experiment results indicate that the proposed method is effective for RV reducer fault diagnosis with the accuracy of 99.69%.

An existing problem in the robotic field is to solve the inverse kinematics (IK) problem of redundant robot with high speed and high precision. A novel IK optimization method based on the Gaussian Damped Least Squares (GDLS) is proposed in this paper. A significant contribution of this method is to make the iteration converge in a faster and more accurate way by introducing an optimal enhanced step-size coefficient. The machine learning model can be trained with 10 ⁶ data points in the reachable region of the robot, and the optimal enhanced step-size coefficient in each solving process can be predicted by the model. The accuracy and stability of the algorithm proposed are verified through an example of an arbitrary 7R redundant robot. The average number of iterations is less than 10, with super high solving speed. Furthermore, the algorithm also has better convergence, which can reach 96.23% when the error threshold is 0.01 mm. The common IK methods are evaluated in this paper, and the results show that the optimized method has good performance in convergence, accuracy and speed.

For the robot system with the uncertain model and unknown environment parameters, a control scheme combining impedance and finite time is proposed. In order to obtain accurate force control performance indirectly by using position tracking, the control scheme is divided into two parts: an outer loop for force impedance control and an inner loop for position tracking control. In the outer loop, in order to eliminate the force tracking error quickly, the impedance control based on force is adopted; when the robot contacts with the environment, the satisfactory force tracking performance can be obtained. In the inner loop, the finite-time control method based on the homogeneous system is used. Through this method, the desired virtual trajectory generated by the outer loop can be tracked, and the contact force tracking performance can be obtained indirectly in the direction of force. This method does not need the dynamics model knowledge of the robot system, thus avoiding the online real-time calculation of the inverse dynamics of the robot. The unknown uncertainty and external interference of the system are obtained online by using the time-delay estimation, and the control process is effectively compensated, so the algorithm is simple, the convergence speed is fast, and the practical application is easy. The theory of finite-time stability is used to prove that the closed-loop system is finite-time stable, and the effectiveness of the algorithm is proved by simulations.

For the 6R robot, there is no analytical solution for some configurations, so it is necessary to analyse inverse kinematics (IK) by the general solution method, which cannot achieve high precision and high speed as the analytical solution. With the expansion of application fields and the complexity of application scenarios, some robots with special configuration have become the research hotspot, and more high-speed and high-precision general algorithms are still being explored and studied. The present paper optimized two general solutions. Elimination is a numerical solution, which has high accuracy, but the solution process is complex and time-consuming. The present paper optimized the elimination method, derived the final matrix expression directly through complex coefficient extraction and simplifying operation, and realized one-step solution. The solving speed was reduced to 15% of the original, and the integrity of the method was supplemented. This paper proposed a new optimization method for the Gaussian damped least-squares method, in which the variable step-size coefficient is introduced and the machine learning method is used for the research. It was proved that, on the basis of guaranteeing the stability of motion, the average number of iterations can be effectively reduced and was only 4-5 times, effectively improving the solving speed.

This paper studies the inverse kinematics of two non-spherical wrist configurations of painting robot. The simplest analytical solution of orthogonal wrist configuration is deduced in this paper for the first time. For the oblique wrist configuration, there is no analytical solution for the configuration. So it is necessary to solve by general method, which cannot achieve high precision and high speed as analytic solution. Two general methods are optimized in this paper. Firstly, the elimination method is optimized to reduce the solving speed to 20% of the original one, and the completeness of the method is supplemented. Based on the Gauss damped least squares method, a new optimization method is proposed to improve the solving speed. The enhanced step length coefficient is introduced to conduct studies with the machine learning correlation method. It has been proved that, on the basis of ensuring the stability of motion, the number of iterations can be effectively reduced and the average number of iterations can be less than 5 times, which can effectively improve the speed of solution. In the simulation and experimental environment, it is verified.