The Research And Implementation Of Process Parameters Optimization And Path Planning For Water-Jet Cutting Robot

Taking Jiangsu Province Scientific and Technological Achievements Transform Project "Flexible High Pressure Water-Jet Cutting System Development and Industrialization" as the background, taking automotive interior part as object, this paper aims to research and develop the cutting parameters optimization method and operation planning system for water-jet cutting robot.A deep analysis and study has been done on the overall structure and related technology of the Flexible High Pressure Water-Jet Cutting System. According to the actual demand, several modules were developed including the prediction cutting quality prediction and process parameters optimization module, the transition path planning module, global path planning module and so on. All of the designing work has finished on the VC++platform, using the Solidworks API combined with OpenGL technology to inplement graphical display and user interaction.A cutting quality prediction model is proposed, based on studying the principle of the 3 layer neural network, the number of hidden layer was increased, and the artificial neural network model of 4 layers is formed to improve the prediction accuracy. Then using the training samples to obtain the relation model between workpiece parameters (feed speed, jet pressure, target distance, particle size of abrasive and abrasive flow rate) and cutting quality(cut surface roughness), then using test samples to validate the prediction effect of the model. Based on the prediction model, gradient method and simulated annealing algorithm respectively was applied to acquire the optimal cutting feed under the desired cutting quality and other constraints.As for transition path planning problem,3 planning schemes are put forward by this paper:1) Gridding automotive interior parts in three dimensional space, and marking the obstacle region and free space region, A* algorithm is proposed to realize the collision free transition path planning; 2) Extracting the features of workpiece surface information, an improved RRT* algorithm is applied to complete collision free trainsition path planning; 3) Based on the improved RRT* algorithm, refer to the idea of bidirectional RRT algorithm, this paper researchs and realizes bidirectional RRT*(Bi-RRT*) algorithm to plan transition path. Finally, select the running time of algorithms and the length of optimization results as indicators, then design simulation experiment to analysis the efficiency between the three algorithms. The experimental results show that Bi-RRT* algorithm is optimal btween the three.Aiming at the problem of previous path planning method, an algorithm combines Bi-RRT* and cutting contour sorting algorithm based on genetic algorithm is proposed to realize the transition path planning at the same time of cutting contour sequence optimization. According to the characteristics of operation planning, designing the connection algorithm of the multi-bidirectional tree, making full use of existing resources to reduce the number of sampling times and generate time of Bi-RRT* algorithm; designing and realizing the transition path database, so that the transition paths between the same adjacent cutting contour do not have to repeated planning, reducing the consumption of path planning and improving the efficiency of planning method. At last, simulation results show effectiveness of the algorithm.Finally, develop the related software and test it to verify the correctness and rationality of this software, then integrat the software to common simulation platform.