| With the rapid development of technology,service robots are gradually widely used in all walks of life.However,at present,service robots are only limited to work on the same floor,and it is difficult to meet the work requirements in high-rise buildings.Therefore,it is of great significance to realize service robots to work across floors autonomously.This paper mainly designs an elevator intelligent key system based on multi-sensor fusion,and proposes a Kalman filtering approach based on complementary filtering and median filtering and an adaptive weighted fusion approach based on Kalman filtering and recursive least squares to implement the system.The function of floor judgment solves the problem that the service robot rides the elevator autonomously.The main research contents are as follows:Firstly,a new type of elevator intelligent key device is designed.The device is mainly composed of auxiliary buttons and a power mechanism,which can not only realize the function of non-contact pressing of elevator buttons by service robots,but also not affect other passengers to manually press elevator buttons manually.It also has an independent floor judgment function.Then,in order to address the problem of inherent drift and noise of the acceleration sensor used for floor judgment,a Kalman filtering approach based on complementary filtering and median filtering is proposed to process the original acceleration value,and the third-order linear Kalman filtering is used to process the pressure sensor data and filtered acceleration Values are processed in real-time data fusion,which improves the accuracy of system floor judgment.At the same time,relevant tests were conducted in the building of Qingdao Ocean Engineering Research Institute of Tianjin University to verify the effectiveness of the method.Finally,an adaptive weighted fusion approach based on Kalman filtering and recursive least squares is proposed to address the problem that the elevator stops operating many times during the process,which increases the height estimation error of the system.By increasing the amount of fusion data and reducing the impact of old data,the problem of height estimation error increase is effectively suppressed.At the same time,relevant test experiments are carried out to verify the effectiveness of the method. |
