Introduction: Urban waterfront spaces are often composed of built infrastructures and nature elements. Though citizens could take advantage of these public spaces to relax from daily work, its restorative potential has not been paid enough attention. In this study, the restorative effect and mechanism of different audio and visual elements in urban waterfront spaces was systematically studied.Methods: At the first stage, restorative potential of waterfront spaces was investigated and different elements with restorative effects were identified through an on-site survey, in which visual and auditory forms of environmental-nature, animal-nature, on-water human activities and on-shore human activities were identified. At the second stage, a series of laboratory experiments were conducted to explore the restorative function of the audio and visual elements.Results and discussion: It is found that the degree of artificiality of waterfront space was a crucial factor influencing the restoration level of the space, and higher artificiality level of waterfront space resulted in lower level of perceived restoration. However it was available by adding visual and audio elements to the scene to facilitate the restorative effect in waterfront spaces with high-level artificiality. The effects of adding visual and auditory elements on psychophysiological restoration were explored, and elements that should be recommended and restrained were discussed.Prospects: These findings would provide applicable suggestions for future design and rebuilding of urban waterfront spaces.

Electric vehicles, known for their low-noise emission, are popular and widespread in metropolises in China, and they provide an opportunity for a reduction in environmental noise from vehicles. To understand the noise from electric vehicles better, this study develops noise emission models considering speed, acceleration, and motion state. The model construction is based on the data collected from a pass-by noise measurement experiment in Guangzhou, China. The models describe a linear relationship between the noise level, the logarithm of speed, and the acceleration for multiple motion states (i.e., the constant-speed state, the acceleration state, and the deceleration state). From the spectrum analysis, the low-frequency noise is barely affected by the speed and acceleration, but the noise at a certain frequency is most sensitive to them. Compared to other models, the proposed ones have the highest accuracy and the greatest ability for extrapolation and generalization.

This paper evaluates the amount of noise reduction when using different pavement materials in two adjacent lanes, where the close-proximity (CPX) method is introduced to analyze the tire/road noise before and after pavement maintenance. We consider four types of pavement materials, including ECA-10, PUC-10, PAC-13, and double-layer porous asphalt pavement (PAC-13+PUC-10), where these materials and their combinations are paved on two adjacent lanes. We measure the tire/road noise with the CPX method using a two-wheel trailer that can install two types of tires in different tests. This study provides some guidelines on controlling traffic noise pollution by using the combination of low noise pavement materials. From the experimental and theoretical results, one can see that the highest amount of noise reduction can be obtained when both the inner and outer lanes use the double-layer porous asphalt pavement. To make a balance between the noise reduction performance and the road maintenance cost, one can have the suboptimal choice, where the inner lane uses PUC-10 and the outer lane uses the double-layer porous asphalt pavement.

Most deep-learning-based multi-channel speech enhancement methods focus on designing a set of beamforming coefficients, to directly filter the low signal-to-noise ratio signals received by microphones, which hinders the performance of these approaches. To handle these problems, this paper designs a causal neural filter that fully exploits the spectro-temporal-spatial information in the beamspace domain. Specifically, multiple beams are designed to steer towards all directions, using a parameterized super-directive beamformer in the first stage. After that, a deep-learning-based filter is learned by, simultaneously, modeling the spectro-temporal-spatial discriminability of the speech and the interference, so as to extract the desired speech, coarsely, in the second stage. Finally, to further suppress the interference components, especially at low frequencies, a residual estimation module is adopted, to refine the output of the second stage. Experimental results demonstrate that the proposed approach outperforms many state-of-the-art (SOTA) multi-channel methods, on the generated multi-channel speech dataset based on the DNS-Challenge dataset.

With the development of urban road traffic, road noise pollution is becoming a public concern. Controlling and reducing the harm caused by traffic noise pollution have been the hot spots of traffic noise management research. The subjective annoyance level of traffic noise has become one of the most important measurements for evaluating road traffic pollution. There are subjective experimental methods and objective prediction methods to assess the annoyance level of traffic noise: the subjective experimental method usually uses social surveys or listening experiments in laboratories to directly assess the subjective annoyance level, which is highly reliable, but often requires a lot of time and effort. The objective method extracts acoustic features and predicts the annoyance level through model mapping. Combining the above two methods, this paper proposes a deep learning model-based objective annoyance evaluation method, which directly constructs the mapping between the noise and annoyance level based on the listening experimental results and realizes the rapid evaluation of the noise annoyance level. The experimental results show that this method has reduced the mean absolute error by 30% more than the regression algorithm and neural network, while its performance is insufficient in the annoyance interval where samples are lacking. To solve this problem, the algorithm adopts transfer learning to further improve the robustness with a 30% mean absolute error reduction and a 5% improvement in the correlation coefficient between the true results and predicted results. Although the model trained on college students’ data has some limitations, it is still a useful attempt to apply deep learning to noise assessment.

The transportation industry is one of the largest consumers of fossil fuels and sources of carbon dioxide emission, with highway transportation accounting for more than 70% of the total. In order to promote efficient, clean, diversified, and intelligent use of energy, this study proposes an innovative application technology model for integrating highway transportation and energy to improve the utilization of clean energy and reduce the proportion of fossil energy and carbon emissions in highway transportation. This study addresses the current problems of green energy development in China’s highways, and combines green low-carbon development policies in international transportation with inspirations based on experience. Based on the principles of safety, efficiency, environmental friendliness, wisdom, and economy, this study systematically analyzes the current situation of green energy application in China’s highway transportation. It proposes one core objective, two major roles, and five major security systems of development measures for green energy application China’s highway transportation with the aim of promoting the transformation and upgrade of China’s highway transportation industry. This study can also help to achieve green and low-carbon development of highway transportation and provides a decision-making reference for the ecological civilization construction goals of the industry.