For New Parallel Machine Tool Cnc System

Aiming at two new type parallel machine tools (PMTs), the theory and method of kinematic analysis,NC automatic programming, interpolation algorithm ,post-processor algorithm,error analysis and compensation of PMTs are studied extensively and systematically in this thesis．Kinematic analysis for main feed mechanism of 2TPS+2RPS type PMT is done by establishing kinematic equivalent model, One and two order effect coefficient matrixs for main feed mechanism of 2TPS+2RPS type PMT are derived. Cutter traces are fitted by least square method and NURBS curve, Weight factors are modified and curve fit accuracy is increased.Algorithms for tool path calculation and cutter path interval calculation are derived.Frame model for CAM system of PMTs is established. Based on time equal division and strategy of coarse interpolation combines with fine interpolation,constant feederate interpolation algorithm for new type PMTs is constructed. With optimum treatment of one order Taylor series, the influence of curvature radius to interpolation accuracy is elimated,and interpolate calculation is simplified.The error of interpolation algorithm is analyzed．A special post-processor that can automatically transfer the cutter location data (CL-data) into machine specific NC commands is essential for real machining applications of PMT.The vector relationship expressions are established through analyzing relations among machine tool coordinate system,movable platform coordinate system and workpiece coordinate system.Post processing method of cutter location data including the position and speedcalculation needed for fine interpolation and the position calculation of workpiece platform is presented.Error model for new type PMT is established based on perturbation method,Strategy of error compensation control for new type PMT,which is based on fuzzy neural network,is proposed. Network structure and train algorithm of network model are discussed.Simulation results show that this strategy can reduce machining error effectively.