| With the advent of the intelligent era,unmanned vehicles have been widely used in many kinds of fields.Obstacle avoidance technology in an unknown environment has become one of the core technologies to overcome the bottleneck of intelligent development of unmanned vehicles.In order to achieve more accurate detection of obstacles and control unmanned vehicles to perform corresponding obstacle avoidance behaviors,it is necessary to introduce a multi-sensor system to achieve the purpose of complementary information.Therefore,the multi-sensor information fusion technology has a very important impact on whether the unmanned vehicle can achieve good obstacle avoidance behaviors.In this paper,the self-designed unmanned vehicles is used as a experiment platform,focusing on multi-sensor information fusion technology and obstacle avoidance technology based on fuzzy control algorithm,which has important theoretical significance and strong practical application value for the development of various unmanned mobile platforms.According to the requirements of the obstacle avoidance function of the unmanned vehicles,the mechanical structure design of the unmanned vehicles was carried out.Then the kinematics model of the unmanned vehicles is established,and the control and driving principles applicable to the unmanned vehicles are discussed.The control system with STM32F407 as the core is determined,the electrical components in the system are designed and selected,and the entire vehicle is built and debugged.Based on the mathematical method of multi-sensor information fusion,a new adaptive weighted fusion algorithm is proposed,and MATLAB is used to simulate the algorithm.Analyzing the simulation results and found that the total variance convergence of the adaptive weighted fusion algorithm is stable at about 1×10-3,the effect is stable,but there is a problem of improper weight distribution.By optimizing the iterative process of the algorithm,the problem of weight assignment is solved and the error of the data fusion result is found to be below 0.2%,indicating that the optimized algorithm has better fusion stability and accuracy.Based on the optimized adaptive weighted fusion algorithm,the possible obstacle avoidance environment and obstacle avoidance strategy for unmanned vehicles are analyzed.The obstacle avoidance algorithm is designed according to the principle of fuzzy control,and the corresponding fuzzy and anti-fuzzy rules are formulated.Through the simulation of the unmanned vehicles under a variety of different obstacles,it is found that the unmanned vehicles can avoid the interference of the obstacle,but there are a lot of large-scale steering problems during the movement,indicating that the unmanned vehicle's movement is unstableIn view of the problems in the fuzzy control obstacle avoidance algorithm,the neural network is used to learn and update the function parameters,and the improvement of the fuzzy controller is realized.By analyzing the simulation results:the improved maximum deflection angle is close to 40°,which is about 20%less than before optimization,and the number of turnings is significantly reduced.The sensor ranging and motor control program are designed and tested,and the effectiveness and reliability of the obstacle avoidance algorithm designed in this paper are verified by the unmanned vehicles experiment. |
PDF Full Text Request