Self-Adaptive Algorithm for Warehouse Truck Speed Measurement with Accelerometer

Abstract: Warehouse trucks are now a practical equipment for lifting and moving goods in warehouses. The working situation for warehouse trucks is complicated. Spilling of machine oil or liquid product leads to low friction coefficient of the floor. Warehouse floors are normally made of concrete. Smooth concrete floors easily become slippery with liquid on the surface. With several tons of load on the fork, the braking distance is longer. An Anti-lock Braking System(ABS) for the truck control system is needed concerning the safety of the operator and people surrounding the truck. The warehouse truck braking is realized by controlling the motor torque. In order to control the torque, the truck wheel slip ratio is needed. With an accurate slip ratio, the control system will have accurate information regarding the vehicle status. To calculate the slip ratio, both the vehicle speed and wheel speed is needed. The wheel speed is derived from the motor speed. The vehicle speed is the challenge of this project. For normal passenger vehicles, sensors are installed on the trailer wheel where no slip occurs. In this case, the requirement is to use a lower number of sensors. There are two reasons of this requirement. It reduces the installation complexity and the cost of the control system. According to the literature study, a sensorless ABS system is impossible realize. Without sensor, it is not possible to estimate the vehicle status. One accelerometer is used in this project. An algorithm is designed and implemented in MATLAB. This algorithm solves the problem with zero-point drifting. Additional features are also added to the algorithm. The algorithm is a self-adaptive algorithm. It operates independently of the installation position of the accelerometer, without additional sensors. The result of the experiment with real measurements is very good. Future work on how to rewrite the algorithm from MATLAB programming to C programming is suggested in the end. Methods of implementation of the algorithm are also recommended.