Trajectory optimization for wall-building robots in accordance with nonlinear viscoelastic cement mortar environment

To address the issue of suboptimal masonry quality in wall-building robots operating within the viscoelastic contact environment of cement mortar, a multi-objective trajectory optimization method is proposed based on Kriging surrogate modeling and the Fractal Evolutionary Particle Swarm Optimization (FEPSO) algorithm in this paper. First, orthogonal experimental design is employed to obtain design variable values, with corresponding objective function values derived experimentally. A Kriging surrogate model linking the objective function to design variables is established to overcome the difficulty in constructing a viscoelastic mechanical model for cement mortar. Subsequently, integrating the Kriging surrogate model, a multi-objective trajectory optimization model for bricklaying is developed. The FEPSO algorithm is employed to solve this model, simultaneously enhancing masonry quality while optimizing other performance metrics. The FEPSO algorithm is compared with NSGA-II and MOPSO optimization algorithms, demonstrating its superiority. Then, the TOPSIS algorithm is applied to derive a compromise solution from the Pareto solution set, which is adopted as the optimal masonry scheme. Finally, the optimal masonry scheme is contrasted with the standard door-shaped trajectory planning method. Results indicate that trajectory optimization increased the wall-building robot's efficiency by 23.66%, reduced energy consumption by 29.33%, and improved trajectory smoothness by 90.47%. Concurrently, environmental contact force decreased by 7.03%, and masonry error decreased from 2.57 to 0.14 mm. The proposed method enhances bricklaying quality and provides theoretical references and practical guidance for trajectory planning and construction quality control of similar intelligent construction robots.