Study Of The Control System For The Embedded Computerized Flat Knitting Machine Based On ARM And ECos

Computerized flat knitting machine is a major equipment in textile industry, as a high-techmechatronic product, its technological progress has great relationship with China’s textile industry.However, due to historical reasons and status conditions, compared with the international advancedtechnology, there is still a big gap of computerized flat knitting machine technology betweendomestic and foreign. The gap is mainly reflected in the aspect of the knitting machine’s lower levelintelligence and the slow speed of research in the key technologies. In addition, China’s mostadvanced computerized flat knitting machines still rely on imports, so we must increase the researchand development capabilities to revitalize the national industry. This paper focuses on the controlsystem of computer flat knitting machine, which is the brain of the whole knitting machine. Basedon a comprehensive analysis of current status of research, a competitive control system of computerflat knitting machine is proposed.In this paper, the development and research status of the flat knitting machine is analyzed firstand the meaning of this research is explained then. On the basis of the introduce of the principle andmechanical structures of the computerized flat knitting machine, and the analysis of the controlrequirements of the various components in the machine, a method adopting ARM and eCosoperating system for the construction of the control system is proposed. Compared with thetraditional control system using the Linux operating system, the control system using eCosoperating system proposed in this paper has many advantages. Based on hierarchical concept of thecontrol system, a two-tier distributed control system is designed, it divided into the main controllayer and head control layer, these two layers communicate via the CAN bus. This design overcamethe shortcoming of the complex connection of these two layers and improved the reliability of thecontrol system of the computerized flat knitting machine.In order to improve the security of the system, Freescale’s ARM9-based i.MX258 processor isselected for the main control layer. The hardware platform of the control system of thecomputerized flat knitting machine is designed, including the main control layer and the headcontrol layer. The main control layer includes interfaces such as memory, CAN bus, Ethernet, LCD,touch screen and motor drive circuit, etc. The head control layer includes solenoid drive circuit,needle selector drive circuit, stepper motor drive circuit. In addition, the signal integrity problems inPCB design of core control board is analyzed. Meanwhile, the software of the control system basedon eCos is designed in this paper. The hierarchy, source structure, kernel mechanism of the eCos operating system are discussed and analyzed, and a development environment for eCos to developeCos application is established. The migration process of the eCos embedded operating systembased on the i.MX258 processor is described in detail, and then the RedBoot and eCos kernel aretested. With the design of the device driver of the platform, the software of the control system isdesigned. The software of the main control layer include pattern documents explanation, diskmanagement, system testing, parameter management, weaving control, etc. The software of thehead control layer include CAN communication programs, needle selector drivers, solenoid drivers,stepper motor drivers, etc. The main control layer complete the weaving task together with the headcontrol layer, and they communicate via the CAN bus.Finally, on the basis of the analysis of the fault detection problems of the circuit board of thecomputerized flat knitting machine, a fault detection system of the head control layer’s circuitboards based on a method of fault tree is proposed, and the system has the advantage of batchtesting, so it can improve the production efficiency. The system is designed based on LM3S5R31and MachXO256C. A large number of fault detection experiments of this system were carried out inthe enterprise and the desired results has achieved.