Millimeter Wave Radar Combines Long Short-term Memory and Energy Storage Embedded System for On-street Parking Space Prediction

Abstract

In this study, a millimeter wave radar was applied to detect the parking status and determine the availability of parking spaces. The data can be quickly uploaded to the cloud so that the parking status can be updated in real time. On the basis of cloud data, a long short-term memory (LSTM) model is built to perform deep learning. The LSTM can provide parking status prediction through the data and enable users to reserve parking spaces in advance, which can effectively increase the utilization rate of parking spaces by nearly 50%. The system can be quickly deployed, uses green energy, and is designed with a small portable photovoltaic (PV) energy storage system with programmable charging technology. To power the equipment, two long-term cycle battery packs are also included. When the remaining power of a battery pack is close to the minimum threshold, a programmable charging system activates the battery assembly and discharging mechanism while using the PV energy storage system to charge the unused battery pack. This design has the ability to extend the battery life by a factor of two, monitor the power status through the cloud, effectively alert technicians to replace batteries, and reduce maintenance labor requirements by 50%.