Research On Motion Control Of Omnidirectional Mobile Vehicle Based On Machine Vision

With the introduction of 2025 made in China,Chinese manufacturing industry gradually towards industrialization,informationization and intelligent.Manufacturing production process put forward higher requirements on the mode of transport,the operation precision and production efficiency,to reduce production time and save labor costs,more and more factories began to adopt a series of logistics automation equipment.AGV(Automated Guided Vehicle)has attracted much interests because of its advantages of high degree of automation and strong flexibility.The dynamic control and intelligent navigation research has been a hot research topic in many industry experts.In this thesis,the design of a four-wheel drive car(Four-wheeled Omnidirectional Mobile omnidirectional mobile vehicle,FOMV)is developed,the main research work is as follows:According to the design requirement of FOMV,some analysis and design related programs are finished from the overall structure,control system,machine vision system.Among them the Mecanum wheel is the key research object,a new Mecanum wheel profile design method is proposed,and the influence of different roller shapes on the Mecanum wheel vibration is summarized.Then the FOMV kinematics model of Mecanum wheel considering skid factors is established,a speed compensation algorithm is proposed by establishing a slip rate model.Finally,the correctness of the model are verified by ADAMS simulation.Furthermore,based on the improved adaptive genetic algorithm,the path planning of FOMV is studied.By modeling the workshop environment map,designing the fitness function with the path length and security as the evaluation index,optimizing the initial population generation method and combining the adaptive genetic operator,the optimal path of the FOMV in the workshop is obtained.Finally,through comparative analysis,it is verified that the algorithm not only accelerates the convergence speed but also avoids falling into local optimum.Then,the optimal path planned is used as the guidance path of FOMV in the workshop,which is identified by machine vision.An equidistant line scan method is proposed to detect the centerline of the guidance route,and the linear equations of the guidance route in the world coordinate system are obtained by least-squares fitting and inverse perspective transformation.Thereby distance deviation and direction deviation of the FOMV and the guidance route are obtained.Further,a FOMV path tracking control method considering slipping is proposed.When the deviation is small,the motion compensation algorithm is combined with the optimal control theory to track the guidance path;when the deviation is large,the fuzzy control theory is used.Afterwards,the designed controller is simulated and analyzed.The simulation results show that the maximum distance deviation does not exceed 5cm,and the direction deviation is within 0.25°.Finally,the 0.5m/s and 1m/s speed tracking experiments is carried out based on the built FOMV prototype vehicle.The experimental results show that when considering slipping,the distance deviation amplitude are 5.28 cm and 5.98 cm respectively due to motion speed compensation,which is 28.79% and 29.77% smaller than those without slipping.Those results verifies that the designed fuzzy optimal controller has a good control effect.The above research work provides a certain reference value and engineering reference for the structure design and the intelligent navigation of FOMV.