Powersplit hybrid electric vehicle control with data dependent systems forecasting
The focus of this paper is on the implementation of Data Dependent Systems (DDS) forecasting in to the control algorithm of the 2001 Michigan Tech FutureTruck. The 2001 MTU FutureTruck is a 2000 model year Chevrolet Suburban and utilizes a powersplit transmission, which is similar to the Toyota Prius, for its hybrid conversion. The main source of propulsion comes from a General Motors, all aluminum block, 3.5L V-6. In the FutureTruck, the accessory current is not directly measured, so it must be calculated from the measured motor current, generator current and battery current. Accessory current is defined as the current used by all of the high voltage components such as the power steering and AC compressor, except the primary drive motor. In order for the vehicle to be charge sustaining, the generator must produce the same amount of power consumed by the accessories and the drive motor. This calculation will only indicate what the accessory load was at the previous sample time and not what the accessory load will be at the current sample time. When it comes to control of the vehicle, this creates a lag, and the controls will undershoot or overshoot the desired accessory current, which creates inefficiencies due to excessive power flow into and out of the battery pack. In order to better understand the accessory load, Data Dependent Systems (DDS) modeling was done on accessory current data collected from the test vehicle, and an AR(26) model was concluded to be adequate, based on the residual sum of squares (RSS) and unified auto-correlations. The DDS modeling of the accessory current also led to the forecasting of the accessory loads. This helps keep battery use to a minimum by allowing the generator to create the correct amount of power, at that time step, to operate the accessories. Accessory draw from the batteries and generator overshoot going into the batteries is minimized and therefore the overall efficiency of the vehicle goes up. The vehicle was tested on a 50-mile circuit including city and highway driving and elevation changes. The results from the test vehicle showed a power savings of 892 kJ/hour which improved the fuel economy by 3 mpg over stock. The charge sustainability of the vehicle was also achieved which means the range of the vehicle is only limited by the fuel mileage, similar to a conventional vehicle.
American Society of Mechanical Engineers, Dynamic Systems and Control Division (Publication) DSC
Powersplit hybrid electric vehicle control with data dependent systems forecasting.
American Society of Mechanical Engineers, Dynamic Systems and Control Division (Publication) DSC,
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/11872