Human-machine-environment information fusion and control compensation strategy for large optical mirror processing system

Abstract

Large optical mirrors require an ultra-precise machining equipment, and a high level of surface-forming precision must be achieved. However, optical mirror processing systems (OMPSs) are susceptible to human behaviors, mechanical structural errors, and processing environments. The factors that affect quality include artificially formulated processes, slurry choice, joint friction, force-induced deformation, ambient temperature, and vibration interference. These factors can lead to a decrease in the accuracy of an OMPS. To study the influence of disturbances in the human-machine-environment (HME) on the OMPS, it is necessary to conduct a fusion analysis of the related factors. A parameter analysis is first conducted on the HME factors that influence the accuracy of OMPS. Then, the factors that influence the accuracy most significantly are determined. Subsequently, with the influencing factors as input parameters, and the output forces of the computer-controlled optical surface (CCOS) grinding system as the output parameters, the HME influencing factors are fused through a BP neural network optimized using a genetic algorithm, and the result is compared with that resulting from the original BP neural network fusion. Finally, according to the results of the fusion, environmental control of the processing system is performed, and the feedforward PD control compensation measures are established for the joint friction. An experimental analysis is also conducted to verify the effect of the information fusion and error compensation on the accuracy of the OMPS.