The Vehicle Speed Prediction Method of Tracked Vehicles Based on Driving Condition Identification

Improving the accuracy of vehicle speed prediction is the key to developing a Tracked vehicle power distribution Strategy. This paper proposes a vehicle speed prediction method based on driving condition identification. The K-means clustering algorithm is used offline to classify the conditions into stationary conditions and quickly-changing conditions, and the current driving condition is judged in real time during the driving process of the vehicle. The Markov-chain prediction method is adopted in stationary conditions while Long Short-Term Memory (LSTM) prediction method is adopted in quickly-changing conditions. Under the premise of maintaining a certain prediction accuracy, it can save calculation time and improve real-time performance.