Multi-level system architecture for massive independent equipment data flow

Abstract

The invention discloses multi-level system architecture for a massive independent equipment data flow. The architecture comprises a data collector, a gateway, a data center and an operating platform, wherein the data collector is connected with the gateway which is connected with the data center, the data center is connected with the operating platform, and the operating platform has a cloud computing function. The architecture disclosed by the invention realizes an industrial large data platform based on a PaaS mode and uses Hadoop as a basic architecture. Moreover, in order to support equipment in various shapes, the system architecture based on the collector-gateway-data center-operating platform is designed, so that changes of user demand can be extremely satisfied.

Description

technical field [0001] The invention relates to a device field solution, in particular to a multi-level system architecture for a large number of independent device flow data. Background technique [0002] Modern enterprises have a large number of production equipment. For the centralized management, monitoring and maintenance of equipment, it is necessary to collect, store and process a large amount of equipment data, which is called stream data. Stream data comes from the control system of the equipment, sensors, and generates data based on time series analog and switch values. In this field, the existing technical framework is an industrial real-time database and a customized data collector, which is used to collect field data and support industrial application software. At present, with the development of big data technology and cloud computing technology, the collection and application of equipment flow data is developing from traditional on-site applications to Intern...

AI Technical Summary

Problems solved by technology

The existing solutions are oriented to the on-site LAN, cannot support Internet applications, and can only realize the release of monitoring screens to the Internet, especially for factory production lines. The existing solutions have the following defects:

[0003] 1. It is difficult to support random addition of devices

[0004] 2. Facing the original point, it cannot effectively distinguish the equipment

[0005] 3. When applied to tens of thousands of independent devices, the cost of implementation and development from collection to application is extremely high, and the variability is extremely weak

[0006] 4. For the plant-level scale, it is difficult to effectively expand capacity and performance in the face of a large number of independent equipment data. It can only be achieved by deploying multiple real-time databases. However, multiple real-time databases are not connected to each other, and data consistency cannot be guaranteed sex

Load balancing can only be achieved by manually developing a coordination mechanism

[0007] 5. It is difficult to support Internet business, and data cannot be accessed through the Internet

[0008] 6. A large number of independent devices are difficult to achieve distributed collection, multiple real-time databases are independent of each other, and management is very difficult

[0009] 7. The collection of different independent devices has two modes: on-site collection and cloud collection, and the existing technical solutions are difficult to adapt

[0010] 8. When faced with a large amount of device data, the local processing capacity is limited, and it is difficult to achieve advanced computing requirements

Images