Research On Inspection Path Planning And Profile Error Evaluation Method For Sculptured Surface

In the fields of machinofacture, aerospace and national defense, parts withsculptured surfaces are widely used, the form error of which has significant influenceon the quality and performance. Research on inspecting techniques and form errorevaluating methods with high efficiency and precision has significant theoreticalsense and practical value. Inspection path planning techniques and profile errorevaluating methods for sculptured surface are mainly studied in this paper.Firstly, inspection path planning of sculptured surface was researched, and amethod generating collision-free optimal path was proposed. First of all, distancematrix was structured. Then the measuring sequence of detection points wasoptimized by using genetic algorithm to obtain the initial path. After that collisiondetection was performed by judging whether interference between the movementswept volume and the curved surface occurred or not, if it did, points avoidingobstacles were added and distance matrix was modified. The above steps wererepeated until the shortest and collision-free inspection path was found.Secondly, profile error evaluation of sculptured surface was studied.Mathematical model for form error evaluation was established by means of minimumzone method. First of all, rough matching was used to eliminate systematic residualsso that the measured points set could have roughly the same position and orientationwith the nominal surface. Then the subdivision and sphere approximation method wastaken to calculate the shortest distance between measured points and nominal surface,and six parameters influencing the position and orientation of the points set wereoptimized by using improved simplex method and differential evolution algorithmrespectively to minimize the maximum one of distances from the measured points tothe nominal surface, thus the profile error of sculptured surface was obtained.Finally, the above methods were verified by simulation. Inspection path wasoptimized by using genetic algorithm. Collision was detected and points avoidingobstacles were added by performing secondary development on UG software with thehelp of GRIP language, when the inspection path planning was finished. Thepractically measured points used in the simulation were got by adding random error tothe detection points, and then the profile error was evaluated by combining subdivision and sphere approximation method with improved simplex method anddifferential evolution algorithm respectively. Simulation results showed the methodsgiven in this paper were effective and feasible.