A signal processing method for dynamic weighing system by SSA-LVQ network

For the error problem of the data processing in dynamic weighing system at present, our experiment takes two-axle vehicle and truck as the object. The collected data by singular spectrum analysis (SSA) algorithm is under preprocessing, then the peak values of the front and rear axle, the ascending slope and descending slope on the weigh-bridge, speed and type of vehicles, is used as input of the Learning Vector Quantization (LVQ) network, and immobile weight as the output to simulate. The results show that: the precision with proposed method is higher than traditional digital signal processing method and traditional LVQ network, and the method effectively reduces the measurement error of dynamic weighing system, and can be applied directly to dynamic weighing system.