【摘要】 在汽车起动机的性能测试设备研制中,负载模拟是一项关键技术。为适应某新型综合测试设备具备多功能和通用性的要求,研制了一种气—液比例控制刹车系统作为模拟负载,并为之设计了智能控制器。该控制器将PID算法与神经网络学习技术相结合,通过在线优化PID参数,提高了负载的闭环特性和自适应能力,保障了对多种型号起动机的参数检测快速、准确。该系统已在起动机生产线上实际运行,效果良好。更多还原

【Abstract】 Load-imitating is a key technology in the development of automobile starters performance CAT bench. In order to meet the multi-function and versatile requirements of a new general type, a proportional-tuning pneumatic-hydraulic brake system functioned as imitated load is developed, a corresponding intelligent controller is also designed for it. The controller combines PID algorithm with a neural network which can optimize the PID parameters dynamically on line, this makes the load system gain fine close-loop characteristics and adaptability which guarantees the performance test to various types of starters quickly and accurately. The load system runs stably and well on a starter assembly line.更多还原