以减小机床热误差,提高加工精度为主要目标,设计以S3C2440A处理器与嵌入式Linux操作系统为控制平台,运用BP神经网络建立误差模型的热误差补偿控制器.首先,控制器通过布置在机床关键温度点上的温度传感器采集加工中心的温度信号,该信号经温度采集模块处理后送到CPU处理器计算出温度值.同时,用激光干涉仪检测出机床对应时刻的误差值.BP神经网络模型根据温度值与误差值计算出综合误差补偿值.然后,将计算出的补偿值通过接口传送给CNC控制中心,CNC控制中心做出误差控制指令,修正机床热变形造成的被加工工件的尺寸误差.仿真实验结果表明了补偿效果的可行性.%Aiming at minishing the therrmal error of CNC machines and improving the machining precision, we design the thermal error compensation controller that using S3C2440A processor and embedded linux operating system for the control platform and using BP neural networks to establish the model of error compensation. First, controller acquire the temperature signal of machining center by the temperature sensor that laid on the key points of machines. The temperature signal is transmitted to the CPU for calculating the temperature value after coping with by the temperature acquisition model. At the same time,measuring the error of machine at the corresponding time using laser interferometric measuring equipment. BP neural networks model calculates comprehensive error compensation value according to the temperature value and the error value. Then, transmitting the calculated value to the CNC control center by interface. The CNC control center modify error control instruction in order to modify the dimension error of workpiece that caused by thermal deformation. The simulation result showed the feasibility of the intelligent compensation.
展开▼
