Control And Research Of AGV Based On Simulated Annealing Reverse Parallel Ant Colony Algorithm

The quality control of galvanized steel pipe fittings is an important premise to improve its corrosion resistance and service life,especially in the application of key components in high precision fields.Production links of steel pipe fittings galvanized soaking time,solution concentration ratio,hot immersion temperature and different models of transportation,most rely on manual experience or semi-robotic operation,can not be real-time monitoring to adjust the parameters of each link to make steel galvanized surface in the best state,the quality of finished products is uneven and poor stability.At the same time,the closed and narrow working environment is accompanied by acid and alkali,high temperature and safety production problems,which makes the labor cost high.In this paper,omnidirectional AGV transport car in the automatic process of new pipe fittings galvanizing is taken as the research object to explore the AGV guidance mode,hardware module,motion model and software algorithm control system suitable for the galvanizing environment,so as to realize the path tracking and process control of the car in the process of pipe fittings galvanizing and improve the yield and efficiency of pipe fittings galvanizing production.So as to complete the workshop transformation and upgrading.Specific research work is as follows:In this paper,aiming at the production environment in the working workshop where each link is crowded and the space is narrow,the planning and design of an electromagnetic guidance AGV driven by two diagonal steering wheels is proposed.Compared with the omnidirectional AGV with four steering wheels or Mc Namm wheels,the control is simpler,the input cost is small,and it can meet the requirements of speed and accuracy of pipe fitting transportation in the galvanizing process.The design and selection of main control module,positioning and navigation module,safety device and sensor,driving unit,wireless communication module and power module lay the hardware foundation for the realization of AGV control system.Secondly,in view of the selection of control parameters and behavior prediction of AGV control system under different loads and operating states,the omnidirectional AGV fourwheel model with diagonal double steering wheels is equivalent to the single-wheel motion model analysis of two virtual wheels(both front and rear wheels are steering + drive),which is convenient to deconstruct and calculate the four-wheel AGV motion model.Then a state estimator based on extended Kalman filter is introduced to design a priori model to improve the self-adjusting and rectifying ability of AGV control system.In addition,on the software algorithm of car control system,aiming at the control precision and deviation correcting speed of car tracking self-adjustment,combining the advantages of simulated annealing and ant colony PID algorithms,an omnidirectional AGV control algorithm,which is based on simulated annealing reverse parallel ant colony PID algorithm,is proposed in this paper.The reverse special ant colony with probability jump and annealing rate in simulated annealing was introduced into the conventional ant colony,which expanded the search range of the solution space neighborhood of the traditional ant colony,accelerated the convergence speed of the ant colony algorithm and the rate and accuracy of selecting PID control parameters in the later period,and improved the stability of the car path optimization and process control.Finally,in the relevant Matlab/Simulink simulation test,the system performance indexes of simulated annealing,conventional ant colony PID and reverse parallel ant colony PID proposed in this paper are tested and compared with the path optimization in the random black and white grid.In order to simulate the AGV system algorithm in the process of running PID control parameter selection and self-adjusting ability rate.The results show that the control accuracy and speed of the simulated annealing parallel ant colony algorithm are 28.3% higher than that of the conventional ant colony PID algorithm.At the same time,during the commissioning of the test site,the omnidirectional AGV experimental prototype with different loads was repeatedly carried out comparative experiments on the road sections with different bending degrees,which verified the excellent speed self-adjustment ability,tracking control and stopping accuracy of the car,which met the production requirements of the project working condition environment.