Embedded Files
[Neural network predicts water quality]
[Neural network predicts water quality]
Predict various water quality indicators by LSTM to improve future water quality conditions.
Direct-reading sensors are used to record data in the cloud. Monitoring items include redox value, dissolved oxygen, suspended solids, ammonia nitrogen, nitrate nitrogen, conductivity, chemical oxygen demand, and pH value.
[Vehicle type detection and counting apply traffic pollution estimation]
[Vehicle type detection and counting apply traffic pollution estimation]
YOLOv4 detects 5 types of vehicles, uses tracking algorithms such as DeepSORT to track each type of vehicle and performs perspective conversion to calculate the number and speed of each type of vehicle.
The number and average speed of each vehicle type (passenger cars, small trucks, buses, large trucks and motorcycles) are used to estimate the contribution of each type of traffic pollution.
[Develop smart trash cans based on image recognition and sound recognition]
[Develop smart trash cans based on image recognition and sound recognition]
The AI image object detection technology of the YOLO series has been used to effectively detect objects such as garbage and microorganisms in the environment. After YOLOV7 was launched last year, it even combined the image segmentation function to develop YOLOV7- Segmentation technology can be used for more effective detection, but there are few studies on applying YOLOV7-Segmentation to garbage classification.
[Multiobjective Optimal allocation of electric scooter battery swap battery station using]
[Multiobjective Optimal allocation of electric scooter battery swap battery station using]
The biosymbiosis algorithm is used to optimize the site selection of electric motorcycle battery exchange station facilities to achieve the goal of minimizing costs and distances between users and facilities.
By combining the geographical information system (GIS) to visualize the optimal plan, it is easy to interpret the facility configuration.
[Effluent water quality prediction and prediction model optimization]
[Effluent water quality prediction and prediction model optimization]
Use deep learning models to instantly predict running water quality
Search for the best hyperparameter through heuristic algorithms to enhance the prediction capabilities of deep learning models
[Real-time Water Quality Sensing and System Server]
[Real-time Water Quality Sensing and System Server]
A water quality monitoring system is being implemented in the die-casting factory, involving the installation of various water quality sensors such as SS, ORP, pH, etc. The data collected will be transmitted in real-time to a central server powered by a Raspberry Pi operating system, using communication protocols like MQTT and Modbus. This setup will integrate data sent from various ESP series development boards. The ultimate goals are to establish a dashboard for data visualization, implement real-time alerts, and automate control systems. This will enable the management staff to have a comprehensive understanding of the water quality status in the die-casting factory.
[Gas Separation]
[Gas Separation]
Use thin film reactors to increase hydrogen production rate and capture carbon dioxide for hydrogen energy industries such as power generation and fuel cells, and industries that require hydrogen raw materials
[Develop AI resource sorting machine] 與 雲科大電機系 毛偉龍老師合作
[Develop AI resource sorting machine] 與 雲科大電機系 毛偉龍老師合作
Use inverse kinematics to obtain the motor's rotation angle and drive the arm to pick up garbage for classification.
Use Yolo and other AI deep learning networks to detect and classify junk objects
Page updated
Google Sites
Report abuse