发动机测试中温度数据的融合方法

The Method of Data-fusion for Temperature in the Testing of Engines

由于发动机测试过程中存在着许多不可避免的疏忽误差,如不对数据进行必要的消除误差的处理,将直接影响对发动机运行状态的客观评价,而采用传统的数据处理方法并不能满足越来越高的测试精度要求,提升硬件系统又将大大增加系统成本投入,不适合大多数小型测试系统的开发和使用.本文介绍了在发动机测试实验中,一种用于温度(冷却水温度、机油温度等)数据采集,基于多传感器(N<10)的数据处理方法:采用分布图法剔除疏忽误差并利用算术平均值算法与分批估计相结合的方法对数据进行融合处理,以估算出温度真值.

There are some negligent errors which we can't avoid in the process of engine testing.If we don't take some measures to eliminate the errors,they would influence making the objective judge of engine's behaviour,but if we adopt the traditional method,testing precision can't meet the increasingly high requirement,while upgrading the hardware system will greatly enhance system cost,which can't suit the empoldering and using of minitype testing system.In this article,when the number of sensors is less than 10,a new method of data processing was introduced:use the distributing chart to eliminate negligent errors,then adopt the method of arithmetic equilibrating and batch estimating to fuse the testing data,in order to estimate the true number.