Optimal sensor distribution for multi-station assembly process using chaos-embedded fast-simulated annealing
This paper presents a novel methodology for the allocation of sensors in multi-station assembly processes. It resolves two core issues pertaining to the determination of an optimal number of sensors to be employed and their best locations. To make the traditional approach more effective, the effect of noise on sensor placement is minimized by maximizing the determinant of the Fischer information matrix. A state-space approach is adopted to model the variation propagation pertaining to the transfer of parts in a given multi-station assembly process. Further, the objective function conceived is significant over other contributions with respect to adding the effect of noise coupled with the sensors. Moreover, a new algorithm is developed to optimize the newly formulated objective function. The proposed algorithm combines chaotic sequences with traditional evolutionary fast simulated annealing (EFSA), hence it is termed chaos-embedded fast-simulated annealing (CEFSA). It can find the optimal sensor distribution with the minimum effect of noise in the sensor data. This paper reports on conceptual work, which underpins the research, and also presents details of a numerical example carried out in an industrial context to test the efficacy of the proposed algorithm. Further analysis reveals that the proposed approach obtains optimal distribution of sensors and offers more generic results compared with previously concluded analysis.