针对传统灰狼优化算法位置更新时勘探与开发失衡,收敛速度慢且陷入局部最优的问题,提出一种改进的灰狼算法(balanced grey wolf algorithm based on fitness back learning,BGWO),引入非线性控制参数,增强算法前期勘探能力,加速收敛;在...针对传统灰狼优化算法位置更新时勘探与开发失衡,收敛速度慢且陷入局部最优的问题,提出一种改进的灰狼算法(balanced grey wolf algorithm based on fitness back learning,BGWO),引入非线性控制参数,增强算法前期勘探能力,加速收敛;在种群迭代阶段采用重心反向学习的最优适应度权重更新策略,平衡算法的勘探与开发。16组基准函数测试结果表明,改进后算法能自适应跳出局部最优,在加快算法收敛速度的同时提高全局收敛能力与精度。将BGWO应用于PV型旋风分离器粒级效率GBDT(gradient boosting decision tree)的建模,提高了GBDT的精度,模型相关系数0.980,均方误差0.00079,BGWO-GBDT与GBDT、PSO-GBDT和GWO-GBDT相对比,建模精度和稳定性明显提高,验证了BGWO的有效性。展开更多
For the purpose of improving efficiency and realizing start–stop function, an electric oil pump(EOP) is integrated into an 8-speed automatic transmission(AT). A mathematical model is built to calculate the transmissi...For the purpose of improving efficiency and realizing start–stop function, an electric oil pump(EOP) is integrated into an 8-speed automatic transmission(AT). A mathematical model is built to calculate the transmission power loss and the hydraulic system leakage. Based on this model, a flow-based control strategy is developed for EOP to satisfy the system flow requirement. This control strategy is verified through the forward driving simulation. The results indicate that there is a best combination for the size of mechanical oil pump(MOP) and EOP in terms of minimum energy consumption. In order to get a quick and smooth starting process, control strategies of the EOP and the on-coming clutch are proposed. The test environment on a prototype vehicle is built to verify the feasibility of the integrated EOP and its control strategies. The results show that the selected EOP can satisfy the flow requirement and a quick and smooth starting performance is achieved in the start–stop function. This research has a high value for the forward design of EOP in automatic transmissions with respect to efficiency improvement and start–stop function.展开更多
文摘针对传统灰狼优化算法位置更新时勘探与开发失衡,收敛速度慢且陷入局部最优的问题,提出一种改进的灰狼算法(balanced grey wolf algorithm based on fitness back learning,BGWO),引入非线性控制参数,增强算法前期勘探能力,加速收敛;在种群迭代阶段采用重心反向学习的最优适应度权重更新策略,平衡算法的勘探与开发。16组基准函数测试结果表明,改进后算法能自适应跳出局部最优,在加快算法收敛速度的同时提高全局收敛能力与精度。将BGWO应用于PV型旋风分离器粒级效率GBDT(gradient boosting decision tree)的建模,提高了GBDT的精度,模型相关系数0.980,均方误差0.00079,BGWO-GBDT与GBDT、PSO-GBDT和GWO-GBDT相对比,建模精度和稳定性明显提高,验证了BGWO的有效性。
基金Project(51405010)supported by the National Natural Science Foundation of ChinaProject(2011BAG09B00)supported by the National Science and Technology Support Program of China
文摘For the purpose of improving efficiency and realizing start–stop function, an electric oil pump(EOP) is integrated into an 8-speed automatic transmission(AT). A mathematical model is built to calculate the transmission power loss and the hydraulic system leakage. Based on this model, a flow-based control strategy is developed for EOP to satisfy the system flow requirement. This control strategy is verified through the forward driving simulation. The results indicate that there is a best combination for the size of mechanical oil pump(MOP) and EOP in terms of minimum energy consumption. In order to get a quick and smooth starting process, control strategies of the EOP and the on-coming clutch are proposed. The test environment on a prototype vehicle is built to verify the feasibility of the integrated EOP and its control strategies. The results show that the selected EOP can satisfy the flow requirement and a quick and smooth starting performance is achieved in the start–stop function. This research has a high value for the forward design of EOP in automatic transmissions with respect to efficiency improvement and start–stop function.
