| In recent years,the automobile field has shown a development trend of "electrification and intelligence",and the adaptive cruise system has been greatly developed as an Advanced Driving Assistance System technology.The current adaptive cruise control system,whether it is constant speed cruise control or car-following cruise control,is aimed at improving driving comfort,and there are few researches on cruise control for special purposes(such as cooperative operation).For the adaptive cruise system of the N3(large truck)cleaning vehicle,not only the adaptive cruise problem of the adaptive cruise system vehicle must be solved,but also the cooperative operation problem during the cleaning operation needs to be considered.In response to the above-mentioned problems and the current state of technological development,based on the perceived needs of N3 cleaning vehicle,this paper designs an adaptive cruise system for daily collaborative operations of cleaning vehicle,which aims to effectively improve the efficiency of vehicle and reduce the overall energy consumption of the adaptive cruise system.Based on the analysis of the perception needs of N3 cleaning vehicle,this paper establishes the theoretical system of the perception layer algorithm of the adaptive cruise system of N3 cleaning vehicle.Based on the deep neural network Faster-RCNN and digital image processing technology,object recognition and cover recognition for road cleanliness is realized.Through the fusion of the target recognition algorithm and coverage recognition algorithm to determine the overall scheme of visual perception of road cleanliness.Starting from the perception results,a research on adaptive cruise decision-making planning method based on finite state machine(FSM)was carried out.The state transition mechanism and incentive function were used to realize the typical state of the vehicle and FSM state matching,which determine the reference target for vehicle control at the next moment.Aiming at the decision planning results and the functional requirements of the auto adaptive cruise system,the vehicle longitudinal dynamics theory is combined with the test data of important components.On the basis of meeting the accuracy and simplicity of the model,a distributed modeling method is used to establish the dynamic characteristics Vehicle Dynamics Simulink Model.The fuzzy PID control optimization in the model enhances the robustness of vehicle speed and bodywork actuator control.Finally,the software test scenario is established.Under the conditions of simulating the real test scenario,the performance simulation study of adaptive cruise system under various operating conditions is carried out through joint simulation of Python and Matlab.The result fully reflects the superiority of the adaptive cruise system collaborative work method to the traditional sweeper driving methods.Finally,relying on the laboratory project,the real-time verification of the adaptive cruise system function algorithm was performed from two aspects: the actual vehicle test of the perception layer algorithm and the real vehicle test of the decision control layer algorithm.The experimental results verify the adaptive cruise control researched in this paper The feasibility of the method and lays the foundation for the whole vehicle design analysis of adaptive cruise system. |