A novel DT-based intelligent experiment method for complex industrial products

Abstract

To alleviate the contradiction between multi-disciplinary coupling, high intelligence, and the difficulty in conducting direct experiments for product testing in the industrial field, a new method of intelligent experiment is proposed based on the digital twin. This method exhibits typical characteristics such as the joint integration of knowledge and data and the fusion of virtual and physical elements. The intelligent experiment technology framework suitable for complex industrial product experiment is established through a four-step intelligent experiment transformation method in the industrial field. These steps include a laboratory prototype experiment, a virtual simulation experiment, a hardware-in-the-loop experiment, and a virtual-physical fusion experiment. Aiming to address three typical engineering problems related to position and posture, posture and performance coupling, and micro force testing, three knowledge and data joint driving methods based on Virtual Reality (VR) multidisciplinary fusion calculations are proposed. A multi-agent operation scene is created, which maintains shape, space–time, physical behavior, performance, and macro–micro consistencies. Design and monitoring experiments are completed to achieve the co-evolution and closed-loop feedback between the physical and virtual dimensions. Taking the hydraulic support used in coal mines as an example, a high confidence agent model is established based on the single-machine hydraulic support posture monitoring and loading test bench. Through a virtual and physical cooperative operation mode involving multi-agents and physical models, six types of experiments are designed, including pose, force, and monitoring. Relevant experimental research shows that the requirements of industrial experiments can be met under laboratory conditions. This experiment framework can provide reliable experimental support for the optimization and intelligent monitoring of complex industrial products.