Development Of Edge Computing Node For Ship Structural Health Monitoring

Since our country joined the World Trade Organization at the beginning of the 21 st century,economic and cultural exchanges between China and other countries in the world have become more often than over.As the cornerstone of ensuring rapid economic development,the safety and stability of ship and cargo transportation has a decisive influence on the long-term development of the national economy.Moreover,in the military field,maritime military vessels have also played a key role in safeguarding the integrity of our country’s territorial and sovereignty.Therefore,to ensure the safety of large ships in the sea,it is an indispensable task to monitor the structural condition of the ship.As a new type of sensing technology,fiber Bragg grating has the advantages of high accuracy,small size,strong anti-electromagnetic interference ability,etc.and it has excellent adaptability in a monitoring environment such as a ship hull.Therefore,combining fiber Bragg grating sensors to develop ship structural health monitoring system has become a mainstream solution.At present,the system basically adopts a system architecture of fiber Bragg grating sensor-demodulator-central processing unit.However,when the number of monitoring points increases and the amount of data increases,this architecture will place higher requirements on the computing performance of the central server and the stability of network communication;in addition,some old ships have quite limited monitoring conditions and require wireless ways to realize real-time monitoring of structural status.In response to the above problems,this paper uses Raspberry Pi as the hardware foundation,combined with fiber Bragg grating sensing technology to develop the ship structure health monitoring edge computing node,and has achieved the following results:(1)Designed the ship structure health monitoring system architecture,and successfully migrated the computing functions that originally belonged to the central server to a place close to the data source,realizing rapid and timely data processing,and then sending the key and final data results to the central server or user terminal.The edge computing node functional modules were divided into three layers and four functional modules to realize the collection,processing,storage,and release of monitoring data respectively.The module design adopted a software model of weak coupling and strong cohesion,and established the standardization of data transmission format.(2)Based on the characteristics of the fiber Bragg grating strain sensor,the stress data is calculated at the edge computing node.According to the strain characteristic parameters of the connected fiber Bragg grating strain sensor,strain conversion was performed on the original wavelength data,and the processing function of the fiber Bragg grating temperature sensor is also added.The data is pre-processed at the edge computing node,thereby reducing the amount of data transmitted to the user terminal to the greatest extent,and effectively ensuring the stability of data transmission.(3)The Maria DB database is used to build a data storage module,and the amount of data that needs to be stored is greatly reduced by optimizing the table structure.In extreme cases,the maximum storage space occupied by the monitoring data volume of the computing node for one month was 739.65 MB,which could be guaranteed the need for long-term monitoring of ships.(4)Created a Django-based web service application by using the Django Web framework for both server-side development and the browser-side using mainstream HTML5,CSS,and Java Script technologies.It mainly included the following functions: the online real-time monitoring interface realized the visualization of real-time monitoring data,the average delay from data collection to release was estimated as 0.35 s.This provides a threshold alarm function and the historical data monitoring interface was responsible for historical monitoring data query and for later download by users.The channel configuration interface can be configured for sensor parameters,and the configurable types include temperature and strain;the user management module mainly realizes personalized functions such as user registration,password modification,and user logout.Utilizing the popularity of the Web and multi-device applicability could decrease the threshold for users.