A new solution to the measurement process planning for machine tool assembly based on Kalman filter

This paper introduces a novel approach for measurement process planning in machine tool assembly by applying the observability principle of state space model which is widely used in control engineering. Initially, state space modeling was carried out for describing the variation propagation and accumulation of assembly process. Then, a mathematical explanation of measurement uncertainty accumulation was presented by means of optimal estimation using Kalman filter. Based on this, an analysis algorithm is developed to find an optimal solution with small estimation errors and money and working time costs, so as to provide designers with feasible measurement plans in assembly. When such method is used, a quantitative approach could be offered to the evaluation of uncertainties of measurement process, which is totally different from traditional methods dependent on experiences. The suggested approach was finally applied to the assembly of a horizontal machining center, on which numerical analyses was conducted to validate the proposed method, and some guidance for measurement planning is summarized from this example analysis.