基于改进狮群算法的汽轮机热耗率模型预测

针对汽轮机热消耗率模型难以精准预测的问题,提出一种基于改进的狮群算法和快速学习网综合建模的方法。首先,针对传统狮群算法易早熟收敛以及在迭代后期寻优速度缓慢导致算法陷入局部最优的缺陷,通过引入禁忌搜索、非线性扰动因子以及黄金正弦策略进行改进;其次,对改进后的狮群算法进行数值验证,结果证明其具有更高的收敛精度和收敛速度;最后,采用某热电厂汽轮机的运行数据建立汽轮机热消耗率预测模型,并将改进狮群算法优化的快速学习网对其进行热耗率预测,将实验结果与其他优化策略进行对比验证,实验结果表明,基于改进狮群算法的快速学习网预测模型具有更高的泛化能力,提高了汽轮机热耗率的预测精度。

Prediction of Steam Turbine Heat Consumption Rate Based on Improvement Lion Swarm Optimization

Aiming at the problem that the model of steam turbine heat consumption rate is difficult to be predicted accurately, a method based on the improved lion group algorithm and fast learning network integrated modeling was proposed. Firstly, the traditional lion swarm optimization algorithm is prone to premature convergence and the slow speed of the algorithm in the late iteration leads to the algorithm falling into the local optimal defect. The algorithm was improved by introducing the tabu search, the nonlinear disturbance factor and the golden sine strategy. Secondly, the improved lion swarm algorithm was numerically validated, and the results showed that it had the higher convergence accuracy and the faster convergence speed. Finally, a prediction model of steam turbine heat consumption rate was established based on the operation data of steam turbine in a thermal power plant, and a fast learning network optimized by improved lion swarm optimization algorithm was used to predict the heat consumption rate. The experimental results were compared with other optimization strategies, and the results showed that the fast learning network prediction model based on improved lion swarm algorithm had the higher generalization ability and improved the prediction accuracy of steam turbine heat consumption rate.