Based on the consideration of individual heterogeneity, it is necessary to analyze the characteristics of bus transit travel behaviour of different latent class groups. Firstly, by constructing a latent class model for distinguishing travel groups, the questionnaire was designed from four aspects: household attributes, personal attributes, travel attributes, and suggestions for the development of urban buses. Then, based on the 2,000 valid questionnaires recovered, 10 indicators, such as the total number of people in the household, the total income of the household, the gender of the individual, the age of the individual, and the time spent on walking, etc., were selected as exogenous variables, and a latent categorization was made based on the indexes of the goodness of fit for the travel groups. The results of the study show that: the bus travel group can be divided into three latent classes, with the student group aged 6 to 17, the male worker group aged 23 to 34, and the female worker group aged 35 to 49, accounting for 6.8%, 42.8%, and 50.4% of the total respectively. There are some differences in the travel behaviour of the three groups, and by analyzing the travel characteristics of each group, suggestions are made for the development of urban buses in terms of rationally setting up bus routes and stops, adjusting the operating hours and implementing "just-in-time buses", improving the environment for riding and transferring, and reducing the cost of travelling on buses.

Watermarking technology has been widely used in copyright protection of video, audio, and image. In this paper, a multi-bit grouping audio watermarking method based on a frame-wise framework is proposed, which utilizes the property of phase spectrum distribution. At first, the phase spectrum of each frame is calculated by applying Fourier Transform. In according with the embedded bit number of each frame, the phase bins are segmented into several subsets. A specific phase pattern is constructed in one of these subsets to carry a specific multi-bit watermark. In the process of watermark extraction, a specific phase pattern is identified in each audio frame to extract the watermark information during the watermark embedding. We evaluated the performance of the proposed method and the results showed that the proposed method achieved satisfactory performance.