Application of genetic optimization-based neural network in air leak detection

Air leak detection plays an important role in ensuring the quality and performance of the product.The detection accuracy is influenced by temperature,test pressure,container size,balance time and so on.The gas movement status of the container is complicated and is difficult to be expressed by accurate function,so it is hard to compensate the error directly.Neural network based on genetic optimization method is proposed.The mathematical model of the relationship between the leakage and influence factors is established.Using this model,the test data is processed,the error of the leakage is compensated.The feasibility is verified through experiments on the air leak detection platform based on differential pressure method and simulations.The accuracy of the air leak detection is increased.