Measurement accuracy improvement and area optimization of CMM based on IPSO-SAA

Abstract Coordinate measuring machines (CMMs) are crucial in ensuring dimensional and technological accuracy of parts in fields demanding high-accuracy production, such as aerospace, shipbuilding, automotive, and large-scale equipment manufacturing industries. Uneven error distribution in the measurement space of CMMs significantly affects the accuracy. To achieve the high-precision measurement using CMMs, this paper proposes an optimal measurement area identification method based on the Improved Particle-Swarm Optimization-Simulated Annealing Algorithm (IPSO-SAA). First, the laser tracer multistation measurement technique was used to determine the volumetric error of the planning points in the CMM measurement space. Then, the geometric error was modelled using the least squares method. Building on this, the objective function was constructed considering the uneven distribution of measurement space errors for different measurements and the geometric inaccuracy of the CMM for circular arcs and straight lines. Finally, the IPSO-SAA algorithm was applied to determine the ideal measurement area of the CMM under varying measurement conditions. A multi-station measuring system utilizing a laser tracer and validation platform was developed for empirical evidence. The findings indicated that (1) when the measurand was a straight line with a length of 50 mm, the straightness error range was 1.13 -2.47μm in the CMM planning measurement space. The straightness error of the measured line was 1.13 μm when positioned within the optimized measurement area. The measurement was placed in the optimal measurement area for enhancement, resulting in a 54.3% increase in accuracy. (2) When the measurand was a circular arc, the sphericity error range was 1.8 -11.7μm in the CMM planning measurement space. The sphericity error of the measured arc was 1.8 μm when positioned within the optimized measurement area. The measurement was placed in the optimal measurement area for enhancement, resulting in an 84.6% increase in accuracy.