Abstract:  Aiming at the demands of real time, fast and accurate load estimation for telescopic handlers with harsh working environment and complex operating conditions, the mathematical model and the solution algorithm for the load estimation of telescopic handlers were studied. Firstly, three load estimation schemes were proposed and analyzed by using various sensors assembled in the existing functional modules of telescopic handlers. After comprehensively comparing the advantages and disadvantages of each scheme, the mathematical model of load estimation was determined and established based on dynamics principle. Then, the loads were taken as the state variable of the estimated system, the real time signals such as the pressure of the hydraulic system, the luffing angle of the boom and the telescopic length of the telescopic boom were taken as the measured values, the load calculation formula established based on the dynamic rotation law was taken as the observation processes between the state variable and the measured values, and the Kalman filter algorithm was used to solve the mathematical model. At the same time, in order to solve the problem that the state variables changed in the recursive processes of Kalman filter algorithm and a large number of new measurement data lost the ability to correct the state variables, a load estimation algorithm was proposed based on improved Kalman filter. Finally, the offline tests and online tests of a super long load telescopic handler with different loads were carried out. The results show that the maximum absolute error of the estimation results is less than 91 kg for 454 kg light load, and the average absolute percentage error is less than 3% for 1100 kg, 2268 kg, 3368 kg and 4536 kg heavy loads. The response time of the stable estimation value may be within 1 s, which is completely better than the actual application requirements. And this method has simple model and strong portability, and may be extended to the online load estimation of cranes, grabbers and other construction machinery.

Key words: telescopic handler, load dynamic estimation, Kalman filtering, correction ability

摘要： 针对工作环境恶劣、操作工况复杂的伸缩臂叉车载重实时快速准确估计需求，对伸缩臂叉车的载重估计数学模型及其求解算法进行了研究。首先，利用叉车现有功能模块中已装配的各类传感器，提出并分析了三种载重估计方案，在综合比较各方案的优缺点之后，确定并建立了基于动力学原理的载重估计数学模型；然后，将载重作为估计系统的状态变量，将液压系统压力、臂架变幅角度和伸缩臂伸缩长度等实时信号作为测量值，将基于转动定律建立的载重计算公式作为状态变量与测量值之间的观测方程，运用卡尔曼滤波算法对该数学模型进行求解；同时，为解决卡尔曼滤波算法在递推过程中状态变量发生改变从而导致大量新测量数据对状态变量失去校正能力的问题，提出了一种基于改进卡尔曼滤波的载重估计算法；最后，对某企业超长载重伸缩臂叉车进行了不同载重的离线试验和在线试验。研究结果表明：对于454 kg的轻载荷，该算法的估计结果的最大绝对误差小于91 kg，而对于1100 kg、2268 kg、3368 kg和4536 kg的重载荷，其平均绝对百分比误差小于3%；趋于稳定估计值的响应时间可在1 s之内，完全优于实际应用需求。该方法模型简单、可移植强，可推广应用到起重机、抓料机等其他工程机械载重动态估计。

关键词: 伸缩臂叉车, 载重动态估计, 卡尔曼滤波, 校正能力