【摘要】 针对AHU系统中的多维、异构、非线性等特点,提出了一种基于改进粒子群的高斯过程回归(IPSO-GPR)算法。在传统的PSO算法上引进了混沌序列和交叉策略,使粒子在运动过程中能够自适应的调整搜索方向和速度,并针对局部最优解附近的区域进行了重点搜索,降低PSO陷入局部最优解的风险。使用平方指数和二次有理两种核函数的组合GPR来应对复杂的AHU非线性系统。对比PSO-GPR方法的预测性能,IPSO-GPR模型预测的送风温度误差保持在0.07℉以下,预测效果更加精确。更多还原

【Abstract】 Aiming at the characteristics of multi-dimensional,heterogeneous,and non-linear in AHU system,a Gaussian Process Regression(IPSO-GPR)algorithm based on improved particle swarm was proposed.In the traditional PSO algorithm,chaos sequence and crossover strategy are introduced,so that the particles can adaptively adjust the search direction and speed during the movement process,and focus on the area near the local optimal solution to reduce the risk of PSO falling into the local optimal solution.The combined GPR of square index and quadratic rational kernel function is used to deal with the complex AHU nonlinear system.Compared with the PSO-GPR method,the supply air temperature error predicted by IPSO-GPR model is less than 0.07℉,and this method achieves a more accurate prediction effect than the PSO-GPR model.更多还原