重载并联六维力传感器及静态标定

针对传感器重载小尺寸需求,提出一种具有混合分支的重载并联六维力传感器,分析了其结构特点和测量原理。搭建了重载并联六维力传感器标定系统,为改善维间耦合及制造误差等对测量精度产生的影响,从标定算法及模型优化方面对其进行了研究。分别利用最小二乘法和BP神经网络算法对加载实验数据进行了处理,分析结果表明BP神经网络算法要明显优于最小二乘法,并通过数据随机分组测试验证了结果的正确性。基于BP神经网络,提出了一种基于人工鱼群算法的BP神经网络算法,并采用优化后的BP神经网络标定算法对实验数据进行了计算分析,结果表明优化后的BP神经网络计算结果较好且稳定,不易陷入局部极值。

Parallel Six-Axis Force Sensor with Heavy-load Capacity and Static Calibration

Aim at the requirements of the sensor on heavy load and small size, a hybrid branch parallel six-axis force with havey-load capacity is proposed, the structural characteristics and measuring mechanism are also explained. The calibration system of six-axis force sensor is built, and in order to improve the effect of dimensional coupling and manufacturing error on the measurement accuracy of the sensor, the model optimization of calibration algorithm is studied. The least square method and BP neural network calibration algorithm are respectively used to calibration analyze the loading experimental data, the results show that the BP neural network algorithm is better than the least square method, and the correctness of analysis results is proved by the grouping test of random data. Based on the BP neural network, a BP neural network algorithm based on artificial fish swarm algorithm is proposed, and the calibration data is calculated and analyzed by using the optimized BP neural network algorithm, the results show that the BP neural network algorithm based on artificial fish swarm algorithm is more stable and difficult to fall into local extremum.