Research On Autonomous Navigation Technology Of Indoor Service Robot Based On Lidar

The autonomous navigation technology is the key to the intelligent service robot in addition to its own service function.Compared with the large computation of visual navigation,lidar navigation has smaller computation and better real-time performance.Whether the high-precision lidar navigation can be realized,the environmental feature extraction based on radar point cloud data and the optimal path planning are the core.For this reason,the following research has been carried out:Firstly,in order to solve the problem of incomplete feature extraction and low accuracy in the process of building robot indoor environment map,a feature extraction algorithm based on angle coefficient and differential curvature is designed.The k-means algorithm of differential genetic optimization and the characteristics of angle coefficient are used to realize the classification and recognition of obstacle point cloud data.The straight arc feature extraction of indoor environment is realized by fusing the least square method based on linear stability and the micro arc recurrence method based on differential curvature.The experimental results of indoor straight arc feature extraction show that the line and arc environment feature extraction effect of the algorithm is more similar to the real situation.Secondly,according to the path planning requirements of indoor service robots,an immune genetic algorithm combined greedy and differential strategies is designed.The probability selection mechanism combining antibody affinity and antibody concentration is used to improve the quality of the population.The dynamic greedy crossover operator is used to expand the selection space of random point crossover operator,which improves the diversity of the population.The differential mutation operator is used to disturb the population.The global search ability and local search ability of the algorithm are considered.The test results of low dimension and high dimension functions show that the new immune genetic algorithm has higher optimization accuracy and faster convergence speed compared with other intelligent optimization algorithms.Then,aiming at the shortcomings of genetic path planning,such as low precision,slow convergence speed and unsmooth planning path,an immune genetic path planning algorithm combining greedy and difference strategy is designed.Firstly,the environment map of the robot is built based on the grid method,and the path planning is transformed into the grid map search.Secondly,the initial path population generated by the barrier free path generation method is smoothed and optimized by the insertion operator and deletion operator.Finally,path selection is carried out based on the probability selection mechanism of fusion antibody concentration and affinity.The cross and differential mutation operations are used to operate the path individuals,so as to obtain the optimal path of the environment.Finally,environment feature extraction and robot autonomous navigation experiments are carried out based on the self-developed indoor service robot platform.The experimental results show that the relevant navigation methods designed in this paper can effectively meet the task requirements of indoor service robot navigation.