| In recent years,water accidents caused by natural disasters such as floods or man-made factors are frequent.To save the safety of life and property,people’s demand for efficient and rapid underwater rescue is increasingly emerging.The first task of underwater rescue is underwater search.However,the underwater environment is extremely complicated and unknown,and the traditional manual search method is inefficient and the safety of rescuers is difficult to guarantee.Therefore,the intelligent marine equipment for rapid underwater search is the inevitable trend of future research.At present,autonomous underwater vehicle(AUV)has become the main tool applied to ocean exploration.With its advantages of fast navigation,intelligent perception and deep-water operation,AUV can be well applied to underwater rescue and search.When the rescue AUV is performing underwater search tasks,how to formulate the shortest search path,how to avoid obstacles safely in the environment with complex obstacles,and how to realize accurate path tracking under large disturbance are the core problems.These need to be solved urgently in engineering practice.Based on the 2020 key R&D plan project of "Research and Development of Key Technologies and Equipment for Underwater Life Detection and Search and Rescue",this paper makes an in-depth study on the global path planning,local path planning and path tracking of emergency rescue AUV.The specific research contents include the following:Firstly,an improved GA-PAO bilevel optimization algorithm is proposed to solve the problem of AUV planning the real shortest search path for underwater suspected life targets in the presence of complex obstacles.The advantages and limitations of classical genetic algorithm are briefly analyzed,and adaptive genetic operator is put forward as an improvement strategy to solve the defects of insufficient calculation accuracy and easy premature convergence.For the problem of "local optimum" which is easy to overcome by a single algorithm,an improved GA-PSO bilevel optimization algorithm is designed to further improve the calculation accuracy and get the optimal solution.Then,based on the underwater emergency rescue environment,the simplified environment model is built by grid method,and the global path planning simulation experiment of the emergency rescue AUV is carried out under this model.By comparing with the traditional optimization algorithm,it is verified that the improved GA-PSO bilevel optimization algorithm proposed in this paper can get the real shortest path,that is,it can make the AUV work out the fastest search strategy.Secondly,in the process of searching the target point for emergency rescue AUV,aiming at the obstacle avoidance problem of complex dynamic obstacles under water,this paper proposes a local path planning method based on Dynamic Window Approach,DWA).In view of many defects in the traditional DWA method,considering unknown underwater dynamic obstacles,this paper applies fuzzy strategy to propose an improved DWA local path planning method.This method can dynamically adjust the relevant weight coefficients according to the relative position relationship between the AUV’s own position and dynamic obstacles and the deviation from the pre-planned path,so as to make local adaptive changes on the basis of the established optimal path to realize safe obstacle avoidance of AUV.Finally,through the simulation experiment,it is verified that the improved DWA method can not only avoid obstacles safely,but also make the path optimal and search time shortest,and at the same time,it can closely link the global path planning with the local path planning,which proves the superiority of the local path planning algorithm.Finally,under the conditions of model uncertainty,complex external interference and actuator input saturation,based on the global optimal path obtained by global path planning,the path tracking control with the expected path as a complex curve is studied.Firstly,the adaptive forward-looking distance selection function is designed.Then,aiming at the drift angle problem,a drift angle adaptive law is proposed,which can compensate the drift angle well and improve the accuracy of path tracking control.Then,based on the backstepping method,the heading controller and speed controller are designed.Considering the model uncertainty and external environmental interference,a finite-time interference observer is designed to accurately observe the interference,and a saturation compensator is designed to solve the input saturation problem of the actuator.The effectiveness and accuracy of the proposed control method are verified by path following simulation experiments. |
