An intelligent braking system composed single-pedal and multi-objective optimization neural network braking control strategies for electric vehicle

Abstract

The braking system is significant to improve the total energy efficiency and ensure driving security of electric vehicles. An intelligent braking system (IBS) composed single-pedal and multi-objective optimization neural network braking control strategies is proposed in this paper to improve the energy economy, braking stability and driving intelligence for electric vehicles. The braking operations of drivers are divided into two parts: (1) releasing the accelerator pedal and (2) stepping on brake pedal. In the first braking operation, a single-pedal regenerative braking control strategy (RBCS) of accelerator pedal based on adaptive fuzzy control algorithm is proposed to improve energy recovery and driving intelligence. Simulation results illustrate that the simulation velocities under the control of adaptive single-pedal RBCS can follow several standard test cycles (including US06, UDDS, LA92 and ECE) very well. The braking energy can be recovered effectively with a less usage of brake pedal. In the second braking operation, a neural network (NN) controller for the composite braking system (CBS) is proposed to optimize the energy economy and braking stability at the same time. The control effect is verified by simulation results under NEDC cycles. The hardware-in-loop (HIL) experiments of the IBS are also conducted in this paper. Compared with a parallel braking strategy used in EU260 electric vehicles, the energy economy of IBS is improved by 3.67% than EU260 in 3 NEDC cycles. IBS performs more closely to the I curve in a specific braking condition with a decreasing braking severity. The time ratio of hydraulic braking in IBS is 2.27% less than EU260 with an increasing driving intelligence.