风场作用下的定日镜轻量化设计

摘要

In view of the sensitivity of dependence and computational inefficiency in traditional optimization methods for lightweight design of heliostat, an approximate optimization approach combining response surface model with sequential quadratic programming was proposed to deal with the lightweight design of heliostat which has nonlinear response with the action of wind based on guaranteeing the strength, stiffness and dynamic property of heliostat. Then, static wind load and fluctuating wind frequency were respectively obtained by numerical wind tunnel test and Davenport power spectrum. The Latin hypercube experimental design scheme and the response surface method were combined to construct the second-order response surface models by fitting with 15 design parameters. Furthermore, the response surface models were optimized by using the sequential quadratic programming. The results show that the total mass of heliostat is decreased by 198 kg, lightweight rating is 12.31% under the condition of ensuring the allowable deflection, stress and first-order natural frequency, and the effect of lightweight is remarkable. In addition, the approach is rapid, efficient and reliable, and it has strong engineering practicability.%针对传统优化方法在定日镜轻量化设计中表现出的灵敏度依赖性和计算低效性等缺陷，提出将响应面模型与序列二次规划算法相结合的方法。该方法在保证定日镜强度、刚度和动态性能等要求的基础上，对风场作用下产生非线性响应的定日镜结构进行轻量化设计。通过数值风洞试验和Davenport谱得到作用在定日镜上的静态风荷载和脉动风频率，利用拉丁超立方试验设计得到的空间样本点进行数值仿真，建立各响应值关于15个设计变量的二阶响应面模型，并采用序列二次规划算法进行定日镜轻量化设计。研究结果表明：在保证定日镜满足许用挠度、应力和一阶固有频率的前提下，定日镜总质量减少了198 kg，减重率达12.31%，轻量化效果显著。同时该方法快速、高效、可靠，具有较强的工程实用性。