中文摘要：

为了提高双直线电动机驱动的同步直接进给轴的运动精度，对该类直接进给轴的全行程热误差在线补偿方法进行了研究。分析了双直接进给轴全行程热误差的影响因素，提出一种基于核偏最小二乘法（Kernel partial least squares，KPLS）和模糊逻辑相结合的双直接进给轴全行程热误差的在线补偿方法。应用激光干涉仪测量其热变形量，使用热电偶和红外测温仪测量进给机构关键点的温度，以时间匹配温度和变形量数据建立统计样本，在均匀离散点位置建立热误差KPLS识别模型，通过在线计算得到离散点热误差补偿量，再根据任意位置与离散点的模糊关联程度，综合计算全行程任意位置处热误差补偿量。以此理论为基础，建立补偿决策函数和补偿系统，依据补偿决策函数智能推断补偿值，通过向数控系统发送补偿码实现在线补偿。在自构建的龙门双直线电动机驱动的直接进给轴平台上，进行全行程热误差在线补偿试验研究，结果表明：混合KPLS与模糊逻辑可以有效的对双直接进给轴全行程热误差在线补偿，经过随机测试验证，补偿后的进给精度提高了50％。

英文摘要：

In order to improve the accuracy of the direct feed axis driven by double linear motors, an online compensation method of full-stroke thermal error for dual direct feed axis is proposed. The factors that influence full-stroke thermal error of dual feed axis are analyzed. And then a hybrid KPLS and fuzzy logic method is applied to improvement of the positioning accuracy of the feeding axis. In the method, a laser interferometer is utilized to obtain the axial thermal deformation of some evenly discrete points. Some thermocouples and infrared thermometers are used to measure the temperatures of the mechanism. KPLS is used to establish a thermal error identification model for these discrete points. The prediction magnitude for on-line error compensation is acquired through the real-time calculation. On the basis of the relevance of any position and key positions, the fuzzy logic is applied to the prediction of the full-stroke thermal error. Compensation decision function and compensation system are established based on the theory. Compensation value is intelligent inferred through the decision function. Compensation code is sent to CNC system to realize online compensation. To demonstrate the procedure of the proposed approach, an experiment is conducted on the self-construction gantry double direct feed axis test rig. The results show that the hybrid KPLS and fuzzy logic method effectively reduces the direct feed axis thermally induced error by 50 % with a random test.