The estimation of bus travel time and providing accurate information about bus arrival time to passengers are important to make public transport system more user-friendly and thus enhance its competitiveness among various transportation modes. However, for the system to be effective, the information provided to passengers should be highly reliable. The model and technique used for prediction plays a major role in enhancing the accuracy and reliability of the system. The present study proposes a model based approach for accurate prediction of bus travel times for the development of a real time passenger information system under heterogeneous traffic conditions that exist in India. The proposed model considers the predicted bus travel time as the sum of the median of historical bus travel times, random variations in travel time over time, and a model evolution error. In order to capture the random variations in travel time, a model based approach with Particle filtering technique is used, wherein inputs are obtained using k-NN algorithm. The results obtained from the implementation of the above method are compared with the measured travel time data and the prediction accuracy is quantified using the Mean Absolute Percentage Error (MAPE). The Performance of the proposed method showed a clear improvement in prediction accuracy when compared with an existing model based approach using Kalman filter that was reported to be work well under similar traffic conditions.