Computing Complexity Reduction for Predictive Control of Engine Thermal Management System
2022-01-0239
03/29/2022
- Event
- Content
- This paper presents the design and performance evaluation of a reduced complexity algorithm for a model predictive control (MPC) which is based on our previously published SAE paper (2021-01-0225). That paper presented a MPC design concept and demonstrated improved energy efficiency by enabling engine pre-cooling based on GPS/Navigation data to recognize future vehicle speed limit and road grade in anticipation of high load demand. When compared to conventional control, the predictive control demonstrated considerable energy and fuel savings. However, the predictive control strategy is much more complicated due to its highly coupled nonlinear behavior. Also, the MPC strategies are limited to the computational resources in engine control units. Therefore, to address these challenges, a reduced-complexity MPC controller is developed to achieve control objectives and to be executed on a modern ECU within a computation budget. One of the key control logics in the previously published SAE paper is to estimate radiator coolant temperature from inputs that are physics-based and require high computational load. This complexity reduction is accomplished through, first, from vehicle speed and road grade, we can calculate engine speed and torque. Then we can determine fuel mass flow rate at a given engine torque and engine rotational speed. Afterward, cylinder wall temperature could be estimated by a look-up table which is a function of engine speed and fuel mass flow rate. Next step is to calculate engine coolant flow according to engine speed since the mechanical water pump is belt driven. More importantly, the estimation of radiator coolant flow is achieved by incorporating both table-based and equation-based methods. As a result, memory usage of the controller is reduced by the equation-based approach and a reduction in computation is achieved by the look-up tables. Finally, the key estimated inputs are fed into a plant model to accurately estimate engine and radiator coolant temperatures. The combination of these complexity-reduction techniques and real-world validation demonstrates a methodology for a cost effective way of implementing on a modern ECU by cutting the computational complexity.
- Citation
- Chen, Y., Holmer, J., Lee, J., and Ha, J., "Computing Complexity Reduction for Predictive Control of Engine Thermal Management System," SAE Technical Paper 2022-01-0239, 2022, .