Design Of Electrical Data Acquisition System And Study On Fault Diagnosis In Residential Buildings

Building electrical systems can provide energy, and take the function of distribution power for buildings. It is an integral part of construction engineering. As people demand for work and living environment is getting higher and higher, and the continuous development of information technology, building electrical covered more widely, the content of the system structure is more complicated. Although compared to residential buildings in industrial buildings and other large public and commercial buildings, its electrical system structure is relatively simple, less electricity equipment complexity, but because of its use function, the electric system of residential type building also has its own characteristics and requirements. Once its electrical system failure, will lead to economic loss, disastrous, even safety accidents. So, need to be in fault occurs can rapidly and accurately judge the reason, location and damage degree, to reduce loss, lower maintenance costs. Therefore, the realization of a capable of building electrical fault intelligent diagnosis technology is particularly important.In this paper, according to the characteristics of residential buildings and its electric system are introduced, aiming at the common breakdown and its reasons are analyzed; For building electric test experiment platform for the secondary development, installed lighting subsystem, and related parameters of the experimental platform were measured, its reliability is verified, which laid a foundation for the follow-up work.In addition, in view of the application of the fieldbus technology in building electric data acquisition is also studied. Select several mainstream fieldbus protocols, from the development cycle, generalizes the project cost and difficulty of development, comparison, finally determine the Modbus bus and CAN bus electrical data acquisition system used in construction; Designed and developed based on Modbus bus architecture electric data acquisition system, completed the design of the hardware and software part, the desired effect, realized the construction the basic function of data acquisition system.Finally, as part of the fault diagnosis theory study, we selected the RBF neural network and SVM algorithm compares the validation, under the condition of the same sample to the diagnosis results analyzed; Results show that the SVM algorithm under small sample space still has good recognition effect, and the diagnosis time is shorter, ability to respond quickly, than RBF neural network algorithm is more suitable for the construction of electric fault diagnosis.