Structure Optimization Of Frame Based On Particle Swarm Algorithm And Its Secondary Development

Semi-trailer towing vehicle is a kind of efficient and convenient transport vehicles. With the constant development of Chinese economy, it plays a more and more important role in long-distance freight, so it attaches great importance to transportation industry. Frame as its main bearing parts, which bears all sorts of complex loads from inside and outside of vehicle, so its strength will directly affect the normal work and safety of semi-trailer towing vehicle. In addition, in order to cater to the trend of energy saving and environmental protection, it need to reduce the total quality as much as possible, ensuring that the frame is under the condition of good performance.In order to solve the above problems, taking semi-trailer towing vehicle frame as research object, based on the finite element theory, PCL language and particle swarm optimization algorithm and other basic knowledge, using PATRAN software as a development platform, This paper develop the user interface for its structure optimization module. This paper's main content is divided into three following parts:1. Aiming at the multi-objective optimization problem of semi-trailer towing vehicle frame structure, this paper utilize PCL language which is embedded in PATRAN to execute secondary development of optimization module. With seting up a menu of particle swarm optimization module directly into the interface, the user can easily set the three elements of optimization through the man-machine dialogue interface, so as to realize automation of the optimization algorithm.2. Then the several commonly used kinds of approximate model technology and the key steps of frame structure's finite element model by CAE software are introduced. Due to selecting appropriate sample points which can fully reflect the whole design space domain is the key to construct approximate model, otherwise selecting more sample points is also in vain. In this paper, it utilize Latin hypercube design method to select samples, on this basis, the RBF approximation technique was used to construct the frame's approximate model between the design variables and objective function3. The multi-objective particle swarm optimization(MOPSO) was studied. The algorithm utilize the banker method which based on the pareto dominance relations to construct the non dominated solution set of particle swarm. External files set is used to save non dominated solution set which was sought currently, and using the adaptive grid method to update the external set, can make the algorithm has a good distribution; And the weight coefficient method is introduced to further improve the algorithm of particle swarm algorithm in the choice of individual extreme value, and the real-time variation strategy is introduced to the external set for selection process of the global extreme value, lest particles were trapped in local optimal area. The simulation analysis of frame structures was completed, that verify the feasibility and effectiveness of the algorithm. Lastly, the solution set which was selected from the result of the particle swarm optimization is used to finite element analysis get of method finite element analysis, that validate the feasibility and effectiveness optimization result.