Strojniški vestnik - Journal of Mechanical Engineering

The End-Trajectory Sliding Mode Control Algorithm Design of Hybrid Polishing Robot Based on Nonlinear Disturbance Observer

Bo Sun — 2026-03-16

The five-degrees-of-freedom (5-DOF) hybrid polishing robot is utilized for machining large optical mirrors. Since the existing kinematics control strategy does not meet high-precision control requirements, it is necessary to develop a dynamics controller to improve the operational performance of the robot. To enhance the trajectory control accuracy of the polishing robot’s end-effector, a sliding mode control algorithm based on a nonlinear disturbance observer is proposed. First, the dynamic model, which accounts for joint friction effects, is derived using the Newton-Euler method, and a complete explicit dynamic model is established through parameter substitution. Subsequently, considering the influence of the coupling of inertia parameters of each component in the dynamic model on computational efficiency, the model is simplified while compensating for errors caused by neglected terms via the Whale Optimization Algorithm-Elman (WOA-Elman) algorithm to reconstruct the dynamic model with error compensation terms. Finally, based on an analysis of the reaching law and the design of the nonlinear disturbance observer, the end-effector trajectory sliding mode control algorithm is developed. Simulation and experimental results indicate that the inclusion of an improved reaching term effectively reduces system chattering. Furthermore, the nonlinear disturbance observer is employed to estimate system errors and external disturbances, significantly mitigating error fluctuations during the convergence process and thus validating the robustness and high precision of the trajectory tracking control system for the polishing robot.

Depth-of-cut Errors in Research on Elastic Deformation of Process System in Camshaft High Speed Grinding

Tao Liu — 2026-03-16

To address the issues related to the elastic deformation of non-circular profiles during high-speed grinding, this study proposes a novel mathematical model for predicting the deviation between preset and actual grinding depths in multi-pass operations. The model establishes a correlation between the feed displacement of the high-speed grinding wheel frame and the rotational angle of the camshaft’s non-circular contour. A series of experiments were conducted on a dedicated high-speed grinding platform to examine the influence of grinding depth, number of grinding passes, and grinding wheel speed on the elastic deformation and the dynamic stiffness of the grinding system. The results show that the discrepancy between the theoretical and measured displacements remains within 5.56 %, confirming the accuracy and robustness of the proposed model. Increasing the number of grinding passes significantly reduces feed errors induced by the elastic concession of non-circular profile, with the maximum elastic deformation displacement decreasing markedly from 68.9 % to approximately 1 % of the preset depth after five passes. This study pioneers the incorporation of the elastic concession characteristics of non-circular profiles into grinding deformation analysis, providing both a theoretical basis and practical guidance for compensating elastic deformation in camshaft grinding, thereby effectively improving machining accuracy and process stability.

Evaluation of Measurement Uncertainty in Creep-Based Determination of Viscoelastic Material Functions of Polypropylene

Urban Kotnik — 2026-03-16

Modern numerical models use time-dependent material parameters as input data to simulate the viscoelastic response of polymers. Reliable numerical predictions therefore depend on the accurate determination of these parameters. Understanding the measurement uncertainty associated with their identification is essential for assessing the expected range and reliability of the simulation results. Although creep-based uncertainty analyses have been reported for other materials, uncertainty evaluations for polymers that require the determination of multiple viscoelastic material functions remain scarce, with existing polymer studies primarily relying on relaxation tests. This study experimentally analyzes the viscoelastic behavior of polypropylene at 60 °C through tensile and shear creep tests based on extensional and rotational rheometry. The tensile, shear, and bulk compliance functions were determined together with their corresponding standard and expanded measurement uncertainties in accordance with the JCGM 100:2008 guideline. Type A uncertainties were found to dominate the overall uncertainty, with relative expanded uncertainties of approximately 3 percent for shear compliance and up to 25 percent for bulk compliance. The study identifies the main sources of uncertainty and proposes strategies for their reduction, including increasing the number of measurement repetitions and improving environmental control. Overall, a comprehensive uncertainty evaluation of the creep-based determination of viscoelastic material functions is presented, leading to more reliable input data for numerical simulations.

Optimization of Simulation Parameters for Wet Concrete Particles Based on Response Surface Methodology and PSO-BP-GA Method

Xiaohui Liu — 2026-03-16

To address the challenges of low calibration efficiency and limited accuracy in the discrete element modeling of wet concrete within high-dimensional parameter spaces, this study developed a parameter calibration scheme that integrates experimental design and intelligent algorithms. It achieved efficient and high-precision inverse function optimization for determing contact parameters, thereby providing a robust foundation for related engineering simulations. Specifically, the repose angle of wet concrete was determined to be 32.07° based on the heap experiment. Through the Plackett-Burman (PB) experiment and steepest ascent experiment, the three parameters with the greatest influence on the repose angle of wet concrete and their optimal value ranges were identified. These parameters are static friction (X1), the coefficient of rolling friction (X2), and surface energy (X3). Subsequently, using the Box-Behnken (BB) test, the optimal 17 sets of combined data for these three significant factors were determined. To establish the objective function between the repose angle of wet concrete and its influencing parameters, and to obtain optimal parameter values, the particle swarm optimization (PSO) - back propagation (BP) - genetic algorithm (GA) method (PSO-BP-GA) is adopted. First, 80 % of the 17 sets obtained from the BB test were used as the training samples for the BP neural network (BPNN), while the remaining 20 % served as test samples. Then, the PSO is used to optimize the weights and thresholds within the BPNN. After deriving the objective function, GA was utilized to perform inverse function optimization, targeting repose angle of 32.07°. Finally, the static friction coefficient (X1) between wet concrete particles was determined to be 0.158, the rolling friction coefficient (X2) 0.187, and the surface energy (X3) 1.580 J/m2. With these parameters, five simulations were conducted, yielding an average repose angle of 32.31°. Compared with the actual repose angle, the relative error was 0.748 %.

Friction Compensation and External Force Estimation for Robotic Systems Using a Fuzzy Neural Network Approach

Jun Wan — 2026-03-16

To address inaccurate external force estimation caused by nonlinear friction in robotic systems, this paper proposes a friction compensation and external force estimation method based on an adaptive neuro-fuzzy inference system (ANFIS). The approach integrates Stribeck friction modeling with a Takagi–Sugeno fuzzy inference structure to identify joint friction parameters from measured data. Experimental results show that ANFIS yields lower identification errors and better generalization performance than baseline methods including fuzzy neural networks, particle swarm optimization, and least squares. The implemented feedforward compensation strategy achieves maximum torque errors of 0.263 Nm and 0.184 Nm for the two joints, lower than those obtained by the compared approaches. By incorporating the identified friction model into a generalized momentum observer with median and Butterworth filtering, the proposed method reduces the root mean square error and maximum absolute error by 18.3 % and 27.9 %, respectively, and achieves a coefficient of determination (R²) of 0.994. In collision detection tests, the method identifies impact events with reduced false alarm rates under the same experimental settings, supporting its applicability to high-precision force control in robotic applications.