Energy consumption forecast model of CNC machine tools based on support vector regression optimized by improved artificial hummingbird algorithm

Abstract

With the development of the manufacturing industry, energy consumption is growing rapidly, which makes the energy crisis and environmental problems become more and more serious. CNC machine tools play an essential role and are the primary energy consumption devices in the manufacturing industry. The accurate prediction of machine tool energy consumption can provide support for energy production plans and reduce energy waste. This paper proposes a novel energy consumption prediction model based on support vector regression (SVR) optimized by an improved artificial hummingbird algorithm (IAHA). Firstly, as the artificial hummingbird algorithm (AHA) may easily get trapped in a local optimum, an improved AHA based on chaotic mapping and local backtracking exploitation strategy is proposed. The chaotic mapping is used to initialize individual positions, which is good for maintaining population diversity. The local backtracking exploitation strategy is employed to improve the local optimization ability. The effectiveness and feasibility of the IAHA algorithm have been verified through 12 benchmark functions. Then, the IAHA algorithm is employed to optimize the parameters of the SVR, and the IAHA-SVR energy consumption prediction model is established. Finally, the effectiveness and feasibility of the IAHA-SVR model are verified through a case study.