Be based on genetic algorithm to solder robot method plans to consider
- Time:
- Click:138
- source:MARSTEN CNC Machining
1 problem description receives product line welder because of solder of car white automobile body solder dot is much, solder treatment task is onerous, use much stage robot to work in coordination commonly, will rise with this solder quality and shorten solder basic cycle time. The article is to soldering dot of robot work solder already allocated finished premise to fall, solder to what only station nods solder robot method undertakes planning to beg solving. Reason this problem is OK problem of translate into TSP. TSP problem is to show viatic business sets out from initial drop, the point that passes through all need returns initial drop, how to arrange reasonable line to make the distance that pass the shortest. The method program of spot welding robot shows solder robot to set out from position of solder clamp origin, if the graph is shown 1 times, choose the place through all solder, return position of solder clamp origin next, make distance the shortest. Graph position of origin of 1 solder clamp 2.
2 if nod H={h1 to solder,build mathematical model, h2, h3, ... , a visit of Hn} is T={t1 orderly, t2, t3, ... , tn} , among them Ti ∈ H, i= (1, 2, 3, ... , n) , the dimensional coordinate of Ti correspondence is (Xi, yi, zi) , because this can build following mathematical models: 3, the genetic algorithm of the basic idea standard of genetic algorithm includes 3 main operations: Choice, across and mutation. Its measure describes as follows: 3.
1 generation is initiative kind group, evaluate initiative kind group in every individual fitness value. 3.
Whether does criterion of 2 judgement astringent accord with a condition. If accord with, output corresponding search result; Implement the following measure otherwise. 3.
3 basises fitness size executes choice operation according to certain kind. 3.
Alternate probability Pc carries out 4 basises corresponding alternate operation. 3.
Mutation probability Pm carries out 5 basises corresponding mutation operation. 3.
6 return measure (2) . Nextpage is shown 2 times like the graph. 4, emulate the effectiveness that begs solution to choose a method for test and verify, apply MATLAB7.
0 write genetic and algorithmic program to undertake emulation begging solving. Solder commonly solder check the number is robot distributive 30~50, because this selects the exercise solder check the number of robot of the 30 solder that it is a dot (if solder spot position pursues 3) . The setting is planted group of dimensions for 80, stop algebra for 1000, alternate probability takes a cost for 0.
8, mutation probability takes a cost for 0.
005. Seek solution 10 times randomly, get optimum solution is the 3rd times 819.
4040, if the graph is shown 4 times, if record of optimum solution iteration pursues 5. Graph 4 graphs 55, the last word passes genetic algorithm the program of three-dimensional to spot welding robot method, what beg solution effectively to give spot welding robot is the shortest solder method, the result that and this kind of method seeks solution gives more accord with effective demand, can shorten spot welding machine is artificial basic cycle time, raise manufacturing metre, can raise 5~10s commonly. CNC Milling