| Existing sandblasting and rust removal equipments are difficult to apply to the automatic sandblasting and rust removal of large steel box girders with complex curved surfaces including U-shaped ribs due to its low load-bearing capacity and poor mobility.Therefore,a parallel robot for sandblasting and rust removal of steel box girder is developed based on the Stewart parallel structure by our research group.Surface roughness is an important technical indicator for testing the quality of sandblasting,which directly affects the adhesion of the coating and is related to the anti-corrosion performance of bridge steel box girder.This paper focuses on the problem of spray gun trajectory planning considering the surface roughness constraints of the parallel robot,and mainly includes two difficulties:(1)The dusty environment of sandblasting and rust removal of steel box girder makes it difficult to monitor the surface roughness in real time.So it is difficult to ensure that the surface roughness meets the process requirements through the trajectory planning and operation control of the parallel robot.(2)Existing researches on the spray gun trajectory planning of the sandblasting robot are mostly aimed at simple regular surfaces such as planes and arc surfaces.Steel box girders with U-shaped ribs have complex curved surfaces and are large-scale components,so it is difficult to plan the spray gun trajectory for the steel box girder.In addition,the ant colony algorithm is easy to fall into the local optimum due to the lack of insufficient initialization pheromone when it is used to solve the problem of complex curved surfaces trajectory planning.To this end,a multivariate fitting method is used to establish a predictive model of surface roughness.On this basis,the steel box girder complex surface is divided into multiple simple and regular sub-slices.By designing hybrid the genetic ant colony algorithm to determine the best way to connect the trajectory of each sub-slice.Therefore,a spray gun trajectory planning method of parallel robot for sandblasting and rust removal of steel box girder considering surface roughness constraint is proposed.The main contributions are as follows.(1)Establish a predictive model by the multivariate fitting method reflecting the quantitative relationship between surface roughness and the three operating parameters of the parallel robot,i.e.,spray distance,spray angle and spray gun movement.For the difficulty to monitor the surface roughness in real time for sandblasting and rust removal of steel box girder,theoretical analysis of the sandblasting and rust removal mechanism and the actual sandblasting operation of the sandblasting and rust removal parallel robot to determine the main operating parameters that affect the surface roughness of the steel box girder.Furthermore,quantitatively analysis of the relationship between the above-mentioned relevant operating parameters and surface roughness are conducted.Then,a multivariate fitting method is used to establish a predictive model of surface roughness.The surface roughness of the steel box girder can be predicted in real time,so as to ensure that the surface roughness of the steel box girder meets the process requirements by subsequent use of trajectory planning and operation control based on the established predictive model.(2)Propose a spray gun trajectory planning method of parallel robot for sandblasting and rust removal of steel box girder considering surface roughness constraint.Based on the above research,an optimization model is established which takes sandblasting efficiency as the optimization goal,and takes the error between the predicted surface roughness value and the expected value,and the range of sandblasting and rust removal parameters as constraints.The optimal operating parameters,i.e.,spray distance,spray angle and spray gun movement are obtained.The complex curved surface of the steel box girder is divided into multiple simple and regular sub-slices,and an improved bounding box method is designed to generate the spray gun trajectory of the sub-slices based on efficiency optimization.The problem of sub-chip spray gun trajectory combination connection is transformed into an open-loop traveling salesman problem.In order to solve the problem of falling into the local optimum due to the lack of insufficient initialization pheromone,a hybrid genetic ant colony algorithm is designed to determine the best way to connect the trajectory of each sub-chip spray gun.Therefore,the spray gun trajectory of the steel box girder sandblasting and rust removal operation is obtained.(3)For the spray gun trajectory planning of the steel box girder sandblasting and rust removal parallel robot,the effect is obvious due to the surface roughness constraints and sub-slices.The MATLAB simulation experiment is conducted by comparing the spray gun trajectory planning method of parallel robot for sandblasting and rust removal of steel box girder considering surface roughness constraint based on the genetic ant colony algorithm with the spray gun trajectory planning method of parallel robot for sandblasting and rust removal of steel box girder considering surface roughness constraint based on the hybrid genetic ant colony algorithm.The simulation results show that the proposed spray gun trajectory planning method owns better optimization results and faster calculation efficiency.For the local steel box girder with U-shaped ribs,the efficiency of sandblasting is compared with that of manual sandblasting.The results show that the sandblasting efficiency can be increased by about 10.08 times based on the proposed method for the spray gun trajectory planning of the steel box girder sandblasting and rust removal parallel robot considering surface roughness constraints. |
