Study On Soft Measurement Of Gasoline Engine Ignition Advance Angle Based On Chaos RBF Neural Network And PSO-SVM

With the increasing demand for automotive products,the rapid development of China's auto industry has become more serious.However,the environmental and energy issues that it brings have become more serious.Therefore,it is necessary to improve the work performance by optimizing the key parameters of the gasoline engine.The ignition advance angle has a great influence on the performance of the engine.Excessive ignition advance angle is likely to cause knocking,and the power is deteriorated.If the ignition advance angle is too small,the engine efficiency will be reduced and the fuel consumption will be increased.The traditional ignition advance angle is mainly determined through table look-up interpolation,and the accuracy is relatively low.Only the experimental calibration is required to accurately obtain the ignition advance angle of each working condition,but the experimental calibration is difficult to cover all working conditions.Therefore,the research on the soft measurement of ignition advance angle has certain engineering value and theoretical significance.This paper focuses on the soft measurement of gasoline engine ignition advance angle.Firstly,the important influence of the ignition advance angle on the engine performance is expounded.The ignition advance angle data calibrated by the known experimental bench is used as sample data,and the optimal network weight wi and the network optimality of the RBF neural network through chaotic algorithm are analyzed.The node center vector ci is optimized,and the simulation of the ignition advance angle is simulated using the optimized RBF neural network.In addition,the radial basis function is selected as the kernel function of the support vector machine,and the particle swarm optimization algorithm(PSO)is used to optimize the penalty factor of the support vector machine.The widths of C and kernel functions ? are used,and the ignition advance angle is predicted by PSO-SVM method.Finally,the advantages and disadvantages of single soft measurement methods are analyzed,and the soft measurement method based on chaotic RBF neural network and PSO-SVM variable optimal weighted combination is used to measure the ignition advance.The angle measure simulation and measure results are compared with the measure results of each single measure method in the paper.The results show that the combined measure method based on chaotic RBF neural network and PSO-SVM proposed in this paper is more ideal than the single measure methods,and has certain advantages in the measure of gasoline engine ignition advance angle.