中文摘要：

关节臂式坐标测量机误差源多且复杂,其测量空间的误差存在不确定性,为了准确快速的得到关节臂式坐标测量机测量空间中的误差,利用标准锥窝对关节臂式坐标测量机进行了空间单点测量精度实验,获得了训练样本和测试样本;利用BP神经网络对空间误差进行了建模,为了提高其收敛速度和运算速度,引入粒子群优化算法（PSO）对BP神经网络模型进行了优化,并对模型进行了预测和验证;结果表明,BP神经网络和PSO-BP神经网络都可以对关节臂式坐标测量机进行空间点误差预测,PSO-BP神经网络模型的预测结果更加精确,相对误差更小.

英文摘要：

The error sources of AACMM were many and complex,and the measurement space error was uncertain.In order to obtain the measurement space error of AACMM accurately,the standard cone was measured on AACMM,and also obtain the training data sample and test data sample,the AACMM＇ s measurement space error model was built up by BP neural network,and the particle swarm optimization algorithm （PSO） was introduced to optimize the convergence speed and the operation speed of BP neural network,and the prediction and verification of the model was carried out.Results show that BP neural network and PSO-BP neural network both can predict the measurement space error,the prediction results of PSO-BP neural network model are more accurate,and the relative error is smaller.