| Summary: | 碩士 === 國立臺灣大學 === 機械工程學研究所 === 107 === In recent years, electric vehicle technology has been developed maturely. The car endurance is still limited by battery capacity. If a proper route planning can be provided to electric vehicle drivers, driving mileage will be increased under the same capacity. Each electric vehicle, with specificpowertrain and motor characteristics, has an optimal energy-efficient speed range. We can find the most energy-efficient route according to the speed range. However, traditional navigation applications or machines only provide the fastest and shortest route options. If electric vehicles drive on a traditional route, it will result in energy waste.
This research analyzes driver’s trip behavior and proposes a two-stage networked real-time energy-efficient strategy for an electric vehicle driven by multiple motors.
First stage is called the networked energy-efficient route planning system. It helps drivers search for an energy-efficient route through the Android navigation APP, implemented in the system. Second stage is called Charging Balancing Strategy. It enhances the vehicle energy performance by torque distribution. Through implementation of the strategy, three goals, which include searching for energy-efficient routes, enhancing endurance, and reducing state of charges between three battery packs, are achieved.
Networked real-time energy-efficient strategy uses Google Maps API as a development tool to obtain route information. Thid research then applies this information to the speed curve model to predict the vehicle speed curve, with velocity and acceleration at each moment. Route energy consumption is calculated by vehicles energy consumption equations, derived from the electric vehicle structure. With the advantage of light computational complexity, it’s applicable for mobile phones and without negative impacts on phones’ operation and electricity. The integrated design of the real-time torque distribution and the charge balance strategy applies particle swarm optimization to real-time distribute motor torque with it’s characteristic of quick response and convergence, so that the motor can be operated in better efficiency overall.
This strategy is verified by model-in-the-loop platform, chassis dynamometer test, and on-road test. Experimental results show that the strategy can identify the energy-efficient route accurately, save 43.2% energy compared with constant proportion torque distribution, and maintained the charge gap between battery packs at 2%
|
|---|
