Improving fuel economy and robustness of an improved ECMS method

Abstract

Hybrid electric vehicles have shown significant improvement for both fuel efficiency and emission reduction, and attracted many researchers. Paramount for the fuel efficiency of HEVs is the energy management control strategies. ECMS (equivalent consumption minimization strategy) is one of the well-known real time power management strategies and has been used extensively in different works; however, its intrinsic difficulty is to find the optimal equivalent factor, which in theory is determined by the a priori knowledge of the complete driving cycle. Different methods have been proposed to solve this issue, but each one has its own pros. and cons. Especially, the applicability of each method for different cycles as well as the computation overhead are two main concerns in the methods presented so far. In this paper, a new method is presented for calculating equivalent factor in the ECMS method. The method does not rely on any prediction nor the a priori knowledge of driving cycles. Its robustness is demonstrated through different driving cycles with distinct characteristics. Our new method will improve the effectiveness and robustness of the ECMS method.

A Binary Controller to Ensure Engine Peak Efficiency for a Parallel Hybrid Electric Car

“…The accumulated driving time is treated as a state variable. Thus, the HEV model is converted into a dynamic model with 3 state variables and 3 control variables in (18)- (20).…”

“…Plenty of control approaches have been developed to solve similar problems. Among those, fuzzy logic [7], [13], and deterministic rule-based methods [14], [15], are widely used in online control due to their low cost on development, [16], [17], Pontryagin's minimum strategy (PMP) [18], [19], and equivalent consumption minimization strategy (ECMS) [20], [21], are typical methods of real-time optimal control and can obtain a performance close to the optimum; however, MPC highly relies on the prediction accuracy while PMP and ECMS require a large effort of "tuning and calibration" to refine co-state or equivalence factor. As an emerging method recently, reinforcement learning [22], [23], does not require accurate models of the dynamic process and the environment.…”

A Supervisory Control Algorithm of Hybrid Electric Vehicle Based on Adaptive Equivalent Consumption Minimization Strategy with Fuzzy PI

“…The relationship between the actual motor torque and the required torque is described as Equation (8). The dynamic response of the torque is considered by adding a first-order inertial link as shown in Equation (9) In addition, to simulate the dynamic process of the engine, a first-order inertial link is added to the output torque of engine as shown in Equation ( (10) where is the actual motor torque after adding a first-order inertial link, m n is the motor speed, m η is the motor efficiency, and b P is the required battery power. …”

A strategy for power management of electric hybrid marine power systems

“…Literature has shown different ways to calculate the equivalence factor, which can be broadly categorized into 4 groups [54]. First group derives a constant equivalent factor.…”