基于神经网络的汽车压力开关测试系统

Automotive pressure switch test system based on neural network

研制了一套基于神经网络的汽车压力开关测试系统。系统通过数据采集卡采集数据,运用神经网络BP(Back propagation)算法进行数据处理,得到了更为准确和稳定的数据;由VC++软件实现测试系统的实时显示,对开关的通断状态进行反馈,确定开关导通所需的压力大小。实验结果表明,测试系统的稳定性比以前提高了15%以上,运用神经网络,误差率降低到以前的80%,能够更好地测试汽车开关压力,避免了开关过程中使用过大压力,从而延长了开关寿命。

A test system was developed to measure the pressure switch parameters based on the neural network. The system gathers data through a data acquisition card, processes the gathered data with the back propagation algorithm of the neural network to get the accurate and stable data. The VC＋＋ software was used to realize the real-time display of the test system to feedback the switch state and the necessary pressure to turn on the switch can be aware of. Experiments showed that comparing to the original counterpart, the stability of test system is improved by 15~, the error is reduced to 80~/oo. The system can measure the pressure to turn on the switch accurately, avoids the unnecessary high pressure in the switching process to prolong the lifespan of the switch.