A Structural Optimization Of Spraying Robot Based On The Deformation Errors Control

The soles spraying process is important in the entire shoe manufacturing process.It can determine the durability of shoes. However, up to now, the soles sprayingprocess is mainly carried out by the manual or semi-automatic methods. Therefore,the production efficiency is low. In addition, the poisonous gas out of glues can badlyinjure the workers during the spraying process. It has to a great extent restricted thedevelopment of shoe industry. Therefore, against to the problems of inefficiency andthe poisonous gas out of glues, the author designs an automatic spraying robot. UsingSolidworks as a method, the mechanical structural of spraying robot is be designedand then the genetic algorithms is used to optimize the key components which affectthe accuracy of spraying robot. Details are as follows:1. The three-dimensional modeling technology of Solidworks was used toachieve the overall design of the spraying robot, the mechanical design theory was beused to carry out detailed parameter calculations on the power unit of the movingmechanism and the rotating mechanism of spraying robot, and selected the type of thekey components.2. The theory of the genetic algorithm and its application in the bolt clampingoptimization method was discussed. Proposed to put the bolt layout position as thecoding method of design variables, the finite element simulation and geneticalgorithm were used to determine the optimal number and location of bolts to obtainthe global optimal solution so as to ensure the driving accuracy and service life of theball screw shaft.3. In order to establish the relationship between the structural parameter of thefixing frame which similar to the cantilever and deformation of the fixing frame in thespraying process, the neural network method was studied on the bases of the finiteelement predictioned deformation. The finite element simulation datas were enteredinto BP neural network to train so that it can determine deformation prediction modelof BP neural network under different fixing frame structural parameters.4. In order to obtain the optimal fixing frame structural parameters, the fixingframe structural parameter optimization model was established on the bases of the BP neural network prediction model, the genetic algorithm was used to optimize thestructural parameter so that minimized the maximum deformation of the fixing framein the spraying process, ensured spraying uniformity as well as shoe durability，andsaved the economic cost.