Research On Structural Optimization Method Of Pressure Vessel Key Components Based On Intelligent Algorithm

The development of science and technology has promoted the productivity.In industries related to pressure vessels,in order to meet the demand for increased productivity,pressure vessel products have a tendency to become larger.Obviously,large pressure vessels require higher cost of production,transportation,and installation.Therefore,it is of great practical significance to carry out research on lightweight design of pressure vessels to reduce costs.At present,one of the main strategies to realize the lightweighting of pressure vessel products is to optimize its structure and find a more effective structural optimization method,which has become the focus of current academic research.Therefore,this paper proposes a strategy for structural optimization of key components of pressure vessels using intelligent algorithms.To verify the feasibility of the proposed structural optimization strategy,Structural optimization based on the central nozzle structure of the lower head of the adsorber.The main work of the thesis includes:1.A method for predicting the stresses of key components of pressure vessels is proposed by using BP neural network.This paper uses neural networks to predict several stress outputs for critical components of a pressure vessel.Firstly,the sample points of the design variables are constructed by the uniform design method,and then the two stress values of these sample point models are obtained in the finite element analysis software ANSYS to form a complete training sample.Finally,the trained BP neural network is trained by using training samples.The trained BP neural network can be combined with the pressure vessel stress intensity assessment theory to approximate the stress constraint function in structural optimization.2.The cuckoo search algorithm is used to optimize the structural components of the pressure vessel.By using the cuckoo algorithm’s strong global search ability and simple algorithm structure,the algorithm is constructed in MATLAB and the appropriate algorithm parameters are selected.An approximate model of the stress constraint function is constructed based on the trained BP neural network and stress strength assessment criteria.Then the objective function,the stress constraint function approximation model and the value range of the design variables are composed into a complete optimization mathematical model.The optimized mathematical model is optimized by the constructed cuckoo retrieval algorithm,and the value of the design variable with the minimum objective function is obtained.Finally,the optimization result of the cuckoo search algorithm is compared with the particle swarm algorithm.3.Developed a prototype system for the optimization of key components of pressure vessels.Based on the above-mentioned implementation of the intelligent algorithm to optimize the structure of the lower head center takeover,this paper uses the App Designer programming platform in the latest version of MATLAB to design the functional module and its user interface of the pressure vessel key component structure optimization prototype system.The callback functions of controls and menus are programmed to perform their functions.At present,the system can smoothly optimize the structure of the lower header of the adsorber in this paper,and there are multiple interfaces for structural optimization of other key components of the pressure vessel.