We applied two machine learning models to detect earthquakes from records observed with seismometers temporarily installed on a volcanic island. The two models are based on different principles: one regards seismic wa...
详细信息
We applied two machine learning models to detect earthquakes from records observed with seismometers temporarily installed on a volcanic island. The two models are based on different principles: one regards seismic waveforms as images, using a convolutional neural network (CNN) to determine the first arrival times of Pwaves, S-waves. The other model regards seismic waveforms as series data. The model processes seismic waveforms as data in a specific order of noise, P-wave, and S-wave, similar to natural language. The purpose of this study is to present the results of using machine learning first arrival times identification models with two principles for noisy seismic waveforms, caused by sea waves and strong winds in volcanic islands, and to evaluate the effectiveness of machine learning models for noisy observation records. We created a Confusion Matrix using first arrival times determined by an expert and evaluated the detection performance of these two models using some metrics of the matrix. Additionally, we assessed accuracy of the model-identified first arrival times by generating a frequency distribution of the difference from the expert's detecting time. The study discovered that the model treating data as series had superior detection ability for noisy data compared to the one treating data as images and the accuracy of the first arrival time detection was also better for the series data model too. We compared the results obtained on this island with those obtained at the permanent station, which is considered to have less noise interference, described in Mousavi et al., 2020. It was found that the difference in detection ability between the two models is slight for data obtained at permanent stations with low noise interference, but that the difference in detection ability between the algorithms of the two models is significant in noisy environments.
The phenomenon of road dust during vehicle operation poses threats to visibility and road safety, while also negatively impacting respiratory health and air quality, becoming an environmental concern. In order to addr...
详细信息
ISBN:
(纸本)9798350386851;9798350386844
The phenomenon of road dust during vehicle operation poses threats to visibility and road safety, while also negatively impacting respiratory health and air quality, becoming an environmental concern. In order to address this issue, this study proposes an AI-based street cleaning vehicle system aimed at reducing dust emissions and providing a clean and safe driving environment, while also preventing driver distraction leading to traffic accidents. The system utilizes two cameras on the street cleaning vehicle to capture images of road conditions and surface pollution, which are then input into an embedded system for real-time image analysis. This technology requires classification of 8 categories, and leveraging deep learning technology like YolactEdge, real-time detection of road pollution is achieved. AI automatically adjusts water pressure for road cleaning and categorizes road pollution into four levels, adjusting water pressure accordingly. This method achieves good results in road pollution identification and adjusting the amount of water for road cleaning. Firstly, in terms of road pollution identification, the average accuracy is 99.85%, and the frame rate on the NVIDIA Jetson Xavier NX embedded system is 13 frames per second. Secondly, in adjusting the amount of water for road cleaning, after measurement, it is found to reduce water consumption by over 40% compared to traditional methods. This system achieves efficient water usage by accurately identifying road surface dirt and debris for application in low-speed street cleaning vehicles traveling at 20 km/hr.
The damage of tobacco pests seriously restricts the high quality development of enterprises. The effective pest control is based on accurate and timely monitoring information of the pest, the traditional monitoring pr...
详细信息
ISBN:
(数字)9781728198880
ISBN:
(纸本)9781728198897
The damage of tobacco pests seriously restricts the high quality development of enterprises. The effective pest control is based on accurate and timely monitoring information of the pest, the traditional monitoring process for information of the pest has many problems, such as heavy workload, low efficiency and high error rate. In this paper, an efficient trapping and analysis system is designed, which consists of five modules: tobacco pests trapping, power supply voltage stabilization, image acquisition, wireless communication and imageprocessing, all modules work together to catch tobacco pests automatically, take pictures of the trap board and upload them to the host computer, after imagerecognition, the number of tobacco pests is calculated, the result can guide us to take effective pest control work. The design of the system has improved the degree of automation and accuracy, reduce the manual workload and work efficiency has also been greatly improved.
暂无评论