Design Of Intelligent Monitoring System For Stereo Parking Garage Based On Machine Learning

The three-dimensional parking garage can greatly utilize the space and alleviate the current situation that the parking garage is in short supply.At present,the three-dimensional parking garage technology has been developed more maturely,but the maintenance problem has become increasingly prominent,mainly due to the abnormal working mode of the motor and the unreliability of the structure.Because the abnormalities of the motor and structure are often reflected on the surface,it is difficult for the maintenance personnel to find out,which has become a major problem for the maintenance of the three-dimensional parking garage.In addition,since the volume of the three-dimensional parking garage is large,the recording of related data is also a big problem.As a driving force for driving artificial intelligence,machine learning has been fully developed in recent years.The essence of machine learning is the ability to train new data by training the massive data.It is a way of learning to simulate the human brain.Machine learning is divided into supervised learning and unsupervised learning.Common supervised learning methods include classification and regression.Common unsupervised learning methods include clustering and dimensionality reduction.Choosing the right machine learning algorithm can greatly improve the correctness of the classification.In order to solve various problems in the preservation of the three-dimensional parking garage,this paper designs a three-dimensional parking garage intelligent monitoring system based on machine learning,using vibration transmitter,current transmitter,network camera,network acquisition card and other equipment to build a body.Use Python as a programming language to realize machine learning algorithm.Support vector machine(SVM)and K-nearest neighbor is used to evaluate the security status of the parking garage and identify the information,and use the database to store key data.The main research contents and work of this thesis mainly include the following parts:1)Analyzed the problems existing in the three-dimensional parking garage systematically.Designed the system architecture of the intelligent monitoring system of the three-dimensional parking garage.The realization scheme is formulated based on the feasibility,reliability,practicability and stability.The best programming language for machine learning is optimized through investigation and research.2)Designed the acquisition system of the intelligent parking system of the three-dimensional parking garage.The vibration transmitter,current transmitter and network camera are the main sensors to realize the system's capture of the original data.Using Ethernet as the bus,the effective control and stable transmission of the sensor are realized.The power supply mode of the POE optimizes the architecture of the system.3)Researched the concept and classification of machine learning algorithms.Selected support vector machine and K-nearest machine learning algorithm as the core algorithm of the system.Deeply studied the machine learning algorithms such as gradient descent,Bayesian theory and logistic regression.The mathematical principles in the machine learning algorithm are derived and the algorithm is implemented by programming language.4)Python is used as the programming language to design the software of the stereo parking garage intelligent monitoring system,which realizes the display of key data,completes the machine learning of data,and store key data in the database.The user interface has been improved to make it simple and practical.At the same time,the verification of system reliability and stability was completed with the laboratory as the experimental scene.