Abstract:
The application of a genetic algorithm to solve the inverse kinematics problem of manipulative robots is considered. The basic concepts of the method of finding solutions using a genetic algorithm are defined. A block diagram of a simple genetic algorithm is presented. It is justified to use multiprocessor computing systems (transputers) to calculate genetic operators. This will greatly increase the efficiency of genetic algorithms. Manipulation systems with three and four links are selected as examples. The problem statement consisted in determining the hinge coordinates of an industrial robot by the specified Cartesian coordinates of the position of the center of the tool (TCP – Tool Center Point) installed on its final link. The results obtained confirm the effectiveness of genetic algorithms in solving inverse kinematics problems of industrial (manipulation) robots. Based on graph theory, the genetic programming procedure is defined as a way to find optimal kinematic structures of robot manipulation systems. The use of genetic programming for modification of object models of manipulation robots is shown. The object representation of the dynamic model of manipulation robots is considered. It is noted that the recombination of objects corresponding to mathematical expressions having mechanical meaning requires kinematic correspondence of the objects used. It is proposed to draw up object diagrams using computer programs that automate this process based on the principle of visual programming (Low-code).
Fulltext:
PDF file (639 kB)