Research And Development Of Shaker Control System Based On Image Processin

Re-election is a commonly used ore beneficiation method that involves the separation of minerals based on differences in their density.It holds a significant position in modern ore beneficiation practices.The shaking table,as a commonly employed reelection device,utilizes the asymmetric reciprocating motion of its bed surface and the inclined water flow to separate minerals.Currently,manual adjustment of the ore receiving plate remains the predominant method during shaking table operation,leading to issues such as unstable product quality and high labor intensity.By introducing new technologies such as computer vision and image processing,real-time monitoring of the shaking table ore beneficiation process can be achieved.The collected bed surface image data can be analyzed and processed,enabling the real-time adjustment of the ore receiving plate’s position based on the detected changes in the separation points between different mineral zones on the shaking table.This approach effectively enhances the stability of product quality in shaking table ore beneficiation.To address the current low level of automation in shaking table production,this paper proposes a research approach for a shaking table control system based on image processing.Firstly,images of the mineral zones on the shaking table bed surface were collected under different environmental conditions,and a dataset with annotations was created for training the detection algorithm.A mineral zone separation point detection network was constructed using Py Torch,and the optimal network structure was determined through training and evaluation using the dataset.Additionally,anchor learning,matching strategies,and transfer learning were utilized to optimize the accuracy of the network.Furthermore,a motion control system based on the Python software hub platform and a 5G gateway was designed,achieving seamless integration through serial communication.A staircase control algorithm was adopted,which comprehensively considers the detection results from the target detection module and calculates the necessary control actions to move the ore receiving plate to the separation point of the mineral zone,successfully addressing the issue of precise movement of the ore receiving plate.The hardware setup of this system includes a computer,a camera,and the motion control system.The software hub platform serves as the core of the system,integrating all hardware components and controlling the overall operation of the control system.Finally,experiments were conducted at an industrial site to validate the functionality and completeness of the motion control system and software hub platform.The experimental data were analyzed,confirming the feasibility of the design approach proposed in this paper.The monitoring device of the shaking table automatic ore receiving system can replace manual labor,resulting in more stable ore beneficiation performance indicators and achieving an "unmanned" mode of operation.In conclusion,the research and development of a shaking table control system based on image processing has achieved the automation of the ore receiving process,improving the accuracy of shaking table bed surface mineral zone separation point detection and the stability of ore beneficiation performance indicators.This development plays a positive role in the advancement of intelligent ore beneficiation.