Research On The Method Of Weighing Error Compensation For Intelligent Truck Scale

A truck scale is an indispensable apparatus of measurement in life and the accuracy of measurement affects the economic interests.With the influence of non-linearity of the load cell,deformation of the Scale's body,etc.There are nonlinear error and eccentric error in truck scale system.The traditional method of error compensation,the adjustment process complicated,poor compensation effect.In order to solve these problems,this paper apply the truck scale's prior knowledge to construct the constraints,a weighing scale method based on weighted smooth constrained neural network is proposed,and a weighing error compensation model is established to improve the weighing accuracy;through a test of truck scale experiment platform which is built by low-power MCU of MSP430F449 as the core processor,According to " JJG555-1996 non-automatic scale generic test procedures";prove the feasibility and effectiveness of this method.This paper design the intelligent truck scale include: signal process module,Single chip minimum system module,AD module,Interface circuit module and power module,etc.When the truck scale is running,first,load the load on the carrier,the weight of the load acts on load sensors through the carrier and the elastomer inside load sensors are deformed,the resistance strain gauge attached to their surface produces a voltage signal proportional to the load.The voltage signal is amplified and filtered by the signal process circuit,and then sent it to the digital to analog conversion circuit.The processor performs the correlation operation on the digital signal to obtain the weighing result.At the same time,in order to reduce the eccentric error and the nonlinear error of the truck scale,this paper establishes a neural network weighing error compensation model based on the weighted smoothness constraint and then proves the effectiveness of the model.Under the condition of national standard,there were a test with the truck scale for eccentric error,repeatability error,indication error and discrimination.Maximum eccentric error was 0.07 kg,maximum repeatability error was-0.17 kg,maximum indication error was-0.18 kg,discrimination was 0.2kg,So the main performance indicators have reached the national standard.