A Study On AGV Quasi-stop Control Based On Load Characteristics Of Start-up Process

Automatic guided vehicle(AGV),as a highly intelligent and automated combination,has been widely used in many industries.AGV is not a separate operation but cooperate with other equipment to complete the production or storage tasks in most occasions.So the stability of the docking accuracy of AGV with other equipment is crucial to the success of the operation.At present,most researches on improving the accuracy of docking are by adding mechanical auxiliary devices,secondary deceleration,supplemented by high-precision sensors such as lidar or vision to form closed-loop control.But increasing the mechanical device will make the AGV structure complex and high cost,and the secondary deceleration will affect the overall working efficiency of the AGV.Using lidar and vision sensor means high cost and high expenditure.This article starts from the practical application and analyzes the situation that the AGV docking accuracy fluctuates greatly and is unstable during the variable load operation in the classic application case of the magnetic navigation AGV.On the premise of ensuring cost and efficiency through a series of improvement schemes improves AGV repeated docking accuracy in variable load operation.This paper first describes a classic application case of magnetic navigation AGV and studies the factors that affect the accuracy of AGV quasi-stop in the case.After analysis,three quasi-stop influencing factors are determined which are RFID positioning identification error,load change,wheel skid,in which load change is the main influencing factor which will cause AGV own inertia change and finally affect the AGV parking accuracy.For the error caused by RFID location identification,a method of RFID capacitive proximity switch combined location trigger is proposed and a long-strip capacitive sensor detection block is designed to ensure the stability and accuracy of the trigger.Aiming at the parking error caused by the load change,it is proposed to establish a quasi-stop mathematical model based on the load change which take the load value as the input object of the associated model and calculate the deceleration parameters matching it.The proposal of polyurethane solid wheel plussteel sand to prevent slippage is proposed to stop wheel slippage.Stelasticity of polyurethane can ensure the solid contact between the wheel and the ground and the larger friction coefficient of steel sand reduces the occurrence of slippage.Expand and analyze the mathematical model of quasi-stop.According to the electrical characteristics of the AGV driving motor,the load change will lead to the change of the driving torque of the motor output and the driving torque will cause the fluctuation of the phase current which is located in the bottom layer of the motor three-ring control.The response speed of the phase current to the load change is the fastest.This paper proposes a method to analyze the mathematical relationship between phase current and load indirectly predict the load based on the above characteristics.After that,discussing the sampling point of phase current.The phase current of AGV start-up acceleration stage is sampled and the trend of phase current in AGV start-up stage is analyzed.Finally,the relationship between load and phase current is verified by Simulink simulation.Designing and builting an experimental platform which including current sampling circuit and velocity sampling circuit to collect the phase current AGV the start-up stage under different loads.The phase current-load prediction model is fitted by least square method.The relative error of the model is 3% and the prediction accuracy is good.Taking the S deceleration curve as the deceleration parameter and the acceleration JM is used as the deceleration parameter.To find out the JM under each parameter combination,experiments were carried out on AGV with different quasi-stop initial speed,different quasi-stop distance and different loads.Analyzing the mathematical relationship between JM and quasi-stop initial velocity,quasi-stop distance and load and establishing the mathematical model of multivariate nonlinear correlation by MATLAB.Selecting different experimental parameters to verify the parking accuracy of the model and it is proved that the parking accuracy can be controlled within 10 mm.The repeated positioning accuracy is good.