Research on PHM Architecture of Cloud Test

Abstract

In view of the low utilization rate of the test data in the current cloud test system, the advantages of cloud test in computing and storage are not fully utilized, and a cloud-based fault prediction and health management (PHM) system based on test data is proposed. Firstly, it analyzes the shortcomings of the existing cloud computing PHM architecture, and proposes a cloud test PHM architecture based on test data. This method can effectively improve the utilization of cloud test data and provide theoretical support for the development of cloud test systems.. Introduction With the popularity of cloud computing technology, cloud test has become a new research hotspot in the field of testing in recent years. Cloud test refers to automatic test technology based on cloud computing platform, which shows low hardware cost, strong processing power, large storage space and using convenience and other advantages. Fault prediction and health management (PHM) refers to equipment management solutions that combine test data, expert knowledge, maintenance support information, and artificial intelligence to achieve monitoring, interpretation, prediction, and management of the operational status of the device, which is under test. How to combine the new technologies in these two test areas will be an important research direction, so this article has carried out related research on this aspect. The cloud test solves the information communication problem of the traditional test system, stores a large amount of test data, how to obtain the performance status of the equipment from these test data, and conducts health management and life estimation, thereby establishing a cloud test PHM system, which will be a important research direction of cloud test. Scholars all over the world have also done a lot of research on this aspect. The article [5] designs the PHM system architecture for command and control equipment for cloud services, but only gives a theoretical framework and does not involve specific implementation. The article [6] proposes a cloud computing aircraft PHM system, but only deals with virtual resource allocation, and does not involve PHM core content. In view of the low utilization rate of existing cloud test data and the imperfect function of cloud test PHM system, this paper proposes a SOA-based cloud test PHM architecture with test data as the core. Cloud Computing PHM Architecture Improvements and Shortcomings The proposed cloud computing PHM system, compared with the traditional PHM system, mainly improves the resource virtualization and networked services, and fails to effectively utilize the powerful data storage and processing performance of the cloud computing platform. Therefore, it is necessary to establish a cloud test PHM system with test data as the core, with data management and data mining as the core functions, give full play to the performance advantages of the cloud test system, and improve the comprehensive support capability of the cloud test system. 331 Cloud Computing PHM Architecture The traditional PHM system can be divided into six layers according to their functions: data processing layer, condition monitoring layer, fault diagnosis layer, fault prediction layer, decision support layer and expression layer. The cloud computing technology has the following outstanding features: (1) Replacing the test subject in the traditional automatic test system, such as a microprocessor or a computer, with a cloud computing platform; (2) Virtualizing test resources into a shared resource pool as services, which are provided to all users; (3) Provide networked services, solve the problem of information silos in the traditional test systems, and provide the possibility of separating test entities from data processing. Based on the above characteristics of the cloud test system, the existing cloud computing PHM system can be divided into a physical resource layer, a virtual resource layer, a service layer, and an expression layer, as shown in Fig. 1 . Figure 1. Cloud computing PHM architecture. SOA-Based Cloud Test PHM Architecture Test data is the core of the test system. The intelligent management and reuse of test data has become a research hotspot in the field of automatic test. The diagnosis, classification and prediction of faults depend on the efficient processing of test data. Traditional test systems store and process test data in isolation, failing to achieve sharing of test data between different test systems and efficient use of massive test data. In the cloud test system, the cloud computing platform relies on networked services to store a large amount of test data. Therefore, it is necessary to establish a PHM architecture with test data as the core, according to the characteristics of cloud test. This paper proposes a cloud test PHM system based on Service Oriented Architecture (SOA), The architecture links these services by dividing the services which can provide different function, and defining corresponding data interfaces. It is shown in Fig 2. (1) Virtual resource layer: In the cloud test system, the test resource and the data processor are separated by the network. The virtual resource pool built by the cloud test system can access standardized test instruments from different regions, different users and different manufacturers, relying on virtualization technology. It is managed and used in a unified manner. (2) Data management layer: The data structure of test data is diverse, the sampling method is diverse, and there are many redundant information, which brings many difficulties to the management of test data. The data management layer includes the capability of standardization, structuring, screening, integration, classification, and storage for the test data.