The turbulence in the transportation of the muck truck and the aging of the sensor, make the data from muck truck sensors contain a lot of noise points. This has a serious impact on the management of construction waste which can lead to This has a serious impact on the management of construction waste which can lead to a great deal of time and money wasted by the regulators and transport drivers. To solve this problem, this paper firstly analyses the fault diagnosis results of vehicle sensors. Based on it, this paper then uses the fuzzy clustering method to creatively build a fault credit system of the muck truck. This fault credit system analyses the past performance of truck sensors and presents the results in a form of reliability. Combining the data with the reliability of the sensors is beneficial to reduce the influence of noise on the discrimination of electronic bills. Finally in the pattern recognition section, this paper improves PSO-ELM method to make the fault credit system of the muck truck can adjust the weight matrix and the offset matrix in the neural network. Therefore, the credit system can directly adjust the result of the electronic single discrimination without wasting extra computing power. The effectiveness and superiority of the method is verified in the dataset collected from real truck.