Engine oil quality analysis based on probabilistic neural network

The engine oil life detection methods generally include chemical detection and optical detection. The former is inconvenient to operate, while the latter has poor accuracy. In order to obtain the life of engine oil by measuring the physical properties such as viscosity, density and dielectric constant of engine oil, the fluid property sensor(FPS) oil sensor can be used to collect the parameters such as viscosity, density and dielectric constant of engine oil, and map them to the actual life of engine oil. The probabilistic neural network(PNN) is used to learn the test samples, forming a nonlinear evaluation mechanism, and establishing an oil expert judgment system to achieve the purpose of independently judging the life of engine oil. After the training of 15 groups of sample data, 7 groups of samples were evaluated and tested in this paper. The test results show that this method can be used for the identification and online monitoring of engine oil quality, which provides a reference for the life judgment and quality replacement of engine oil, conforms to the actual application requirements, and has certain application value.