| In recent years, with the rapid development of the aerospace industry in everywhere of the word, especially in China, the truss structure is applied in the aerospace industry more and more widely, too. However, the truss structure in space environment with potentially dangerous. The air in the outer is very thin, so the truss structure is very difficult to quickly restore stability when motivated by any load. And that is unfavorable to the space structure itself and the equipments mounted on the truss. To solve the problem, we could control the truss actively. We can fix the actuator and the sensor on the structure, so we can monitor and control the truss in any time. Though we can obtain more information and larger force by more actuators and sensors, the too more actuators/sensors and unreasonable position will not only reduce the control effect but also increase the energy consume, even to the extent that will destroy the whole control system. Therefore, we should optimize the number and layout of actuators/sensors in the truss to control and monitor the truss by the reasonable number and layout. And that has great theoretical significance and practical application value.Against the defect of genetic algorithm in convergence problems for the genetic algorithm, this paper improved the genetic algorithm. And while the genetic algorithm for solving multi-objective when using the traditional fixed weights to determine the target weight coefficients subjectivity shortcomings, this paper using the method of random weights to improved the problem. At the same time, this paper using the APDL of ANSYS to mode and solve the space truss, and part of the command stream is given.Finally, based on the efficiency of actuating, this paper use the improved genetic algorithm to optimize the number and location of the sensors and actuators in truss. After analysis the optimization results, we can know the optimization results is right, so the proposed method has better optimization results. |
PDF Full Text Request