Measurement-based two-layer optimization scheme for machining system with repetitive and batch characteristics
Jing Wang,Jie Zhang,Fuwang Zhang,Zhongbo Sun,Zhen Sui,Yantao Tian
DOI: https://doi.org/10.1007/s00170-022-09460-1
IF: 3.563
2022-06-13
The International Journal of Advanced Manufacturing Technology
Abstract:Most machining systems repeat predetermined tasks during machining. In most cases, the measurement of parts is carried out after the production of parts. However, in the actual production process, even if the system measurement is allowed or controllable in the whole production cycle, it often needs to pay more costs and the implementation is more complex. With the intention of improving the above characteristics of the industrial process with repeated batch processing and the optimization problem of the machining system that cannot be measured frequently, based on this, this paper proposes a measurement-based integrated two-level optimization strategy. In this two-layer strategy, the repeated update strategy is used to perform within-run correction to ensure the tracking accuracy and the cycle-to-cycle corrections are based on cycle-to-cycle control adaptive scheme to improve the optimization performance. The scheme combines real-time tracking and offline optimization, and the reactivity of the processing system is effectively improved. In order to make the two-layer optimization scheme more versatile, this paper introduces the periodic time-varying system and increases the flexibility of constraints and applies it to CNC cam grinding. Simulation and experiments verify the effectiveness and superiority of the scheme.
engineering, manufacturing,automation & control systems