Simultaneous Optimization of Supervisory Control and Gear Shift Logic for a Parallel Hydraulic Hybrid Refuse Truck Using Stochastic Dynamic Programming

Abstract

The power management controller of a hybrid vehicle orchestrates the operation of onboard energy sources, namely engine and auxiliary power source with the goal of maximizing performance objectives such as the fuel economy. The paper focuses on optimization of the power management strategy of the refuse truck with parallel hydraulic hybrid powertrain. The high power density of hydraulic components and high charging/discharging efficiency of accumulator with no power constraint make hydraulic hybrid an excellent choice for heavy-duty stop and go application. Two power management strategies for a parallel hydraulic hybrid refuse truck are compared; heuristic and stochastic dynamic programming based optimal controller. For designing a SDP based controller, an infinite horizon problem is setup with power demand from driver modeled as random Markov process. The objective is to maximize system level efficiency by optimizing (i) the power split between engine and hydraulic propulsion unit, and (ii) gear shift schedule. This combines the optimization of powertrain parameters with power management design.