Fault diagnosis method of sensors in building structural health monitoring system based on communication load optimization

Abstract

The building structure maintenance and safety monitoring system has become an important guarantee of building structure safety. The service life of conventional large-scale buildings is usually fixed in hundreds of years, while the sensor life of the corresponding structural health monitoring (SHM) system can only be maintained in more than ten years or even shorter. Therefore, it is very important and significant to identify and detect the sensor fault of the building SHM system in time and effectively. Based on the communication load optimization technology, this paper will control and optimize the communication load and energy efficiency of a large number of sensor devices, so that the whole monitoring system network has the advantages of small flow and large amount of connected data. At the same time, according to the generalized quasi natural analogy test principle, a sensor fault self diagnosis method is proposed, so as to further quickly realize the detection system sensor fault and fault channel determination. Based on this, the sensor fault detection algorithm of the communication load optimization based building SHM system proposed in this paper is applied to the structure safety monitoring of a large building. The experimental results show that the diagnosis results of this method are accurate and consistent with the actual situation.

Introduction

With the continuous development of the global civil engineering and related technologies, the world’s large-scale buildings and complex buildings are increasing progressively. The large-scale buildings are faced with a large number of problems, such as huge scale, complex structure form, and complex use environment. Once these buildings fail or serious human accidents occur, huge economic losses and casualties will be caused [1], [2], [3], [4], [5]. Therefore, in order to further predict, diagnose and repair large-scale building faults in advance, building SHM system based on a large number of advanced sensors has become an important technology to evaluate the health of buildings [6], [7], [8]. The building SHM system mainly relies on the actual measurement information processing of the structure environment, load and response to realize the evaluation of the current building safety technology status, so as to provide human with the early warning of major accidents and structural safety [9]. While the effectiveness and safety of the building SHM system are completely dependent on a large number of sensors, however, the traditional sensor detection technology completely depends on manual self inspection, which has serious disadvantages such as high cost and poor timeliness [10], [11], [12], [13]. Therefore, how to achieve effective and timely sensor fault detection of building SHM system has become the top priority of civil engineering and construction industry.

Based on the above analysis, a large number of civil engineering practitioners and related research institutions have conducted in-depth research and Discussion on sensor fault detection technology. The mainstream sensor fault detection technology is mainly put forward by German scientists in three ways: analytical model-based method, signal processing based method and knowledge-based method, which establishes the basic theoretical model of sensor fault detection technology [14], [15], [16]. In terms of specific algorithm research, American scientists [17], [18] have proposed sensor fault detection technology based on artificial neural network technology, which has strong advantages such as high fault tolerance, fast response, strong learning ability, adaptive ability and nonlinear approximation ability, but it is too complex and prone to failure; European scholars [19], [20], [21] have proposed based on fuzzy inference The method of theory realizes the detection of sensor fault, which emphasizes the use of strong structural knowledge expression ability of fuzzy logic to realize the processing of uncertain information and incomplete information, but in essence, this method has problems such as knowledge and information acquisition difficulties, and the fuzzy relationship between the corresponding sensor fault and symptom is relatively difficult to determine.

Based on the above analysis, in order to further solve the problem of sensor fault detection in the current building SHM system, this paper will control and optimize the communication load and energy efficiency of a large number of sensor equipment based on the communication load optimization technology, so that the whole SHM system network has the advantages of small flow and large amount of connected data, and at the same time, aiming at the generalized analogy test principle, The method of sensor fault self diagnosis is put forward, so as to realize the judgment of sensor fault and fault channel in the system. Based on this, the sensor fault detection algorithm of the communication load optimization based building SHM system proposed in this paper is applied to the structural safety monitoring of a large building. The experimental results show that the diagnosis results of this method are accurate and consistent with the actual situation.

The structure of this paper is as follows: In the second section of this paper, the communication load optimization technology and the sensor self diagnosis technology based on this technology will be analyzed and studied in detail, which will be the theoretical basis for the third section. In the third section of this paper, a large-scale building in a city will be measured to verify the consistency between the diagnosis results and the actual results.