Information complex diagnostic system for hybrid and electric vehicles

Abstract

Artificial neural networks in the control system of the power plant of the vehicle are considered in order to reduce energy consumption and diagnose the off-line technical condition of the traction battery. A method for diagnosing the technical state of a power plant has been obtained, which uses artificial neural networks and fuzzy inference systems to determine the technical state of an internal combustion engine and a traction battery. The aim of the work is to increase the efficiency of diagnostics of functional systems of a hybrid and an electric vehicle by prompt synthesis of control actions according to energy and quality criteria, taking into account external operating conditions. Substantiation of the method for diagnosing the technical condition of the power plant of a hybrid and electric vehicle using an artificial neural network and a fuzzy inference system. Give a scientific basis for the diagnostic parameters of the power plant of a hybrid vehicle. In the work, artificial
neural networks were used in the control system of the power plant of the vehicle in order to reduce energy consumption and diagnose the off-line technical condition of the traction battery. With the help of the simulator, an unmeasured car model is learned, which uses off-line training of the neurocontroller. The quality of the neurocontroller training is determined by the simulator. With the further functioning of the control system, the parameters of the neural networks do not change. The lack of adaptation of weight coefficients during the operation of the control system is justified by the fact that this leads to the loss of long-term memory of the control system when a short-term malfunction occurs, as well as the possibility of bifurcation during adaptation in nonlinear systems are shown in Figure 1. The target function of optimization of control implies minimization of energy consumption while maintaining the state of charge of the traction battery with a limited range of motion of the vehicle under given operating conditions.
Based on the results of tests of the neurocontroller method, it was found that the neurocontroller provides a decrease in fuel consumption by 17% and reduces the range of changes in the degree of charge of the traction battery by 35%, and also provides minimization of toxic emissions.