Subset simulation for optimal sensors positioning based on value of information

Abstract

Greedy and non-greedy optimization methods have been proposed for maximizing the Value of Information (VoI) for equipment health monitoring by optimal sensors positioning. These methods provide good solutions, but still with limitations and challenges: greedy optimization does not guarantee to find the optimal solution, due to the non-submodularity of the VoI; non-greedy optimization does not suffer from the non-submodularity of the VoI but requires computationally expensive and tedious simulations to find the optimal solution. In this work, the Subset Simulation (SS) method is originally proposed to address these limitations and challenges. A real case study is considered concerning the condition monitoring of a Steam Generator (SG) of a Prototype Fast Breeder Reactor (PFBR). Results show that SS, even if initialized with a small number of Monte Carlo samples, is capable of finding the optimal set of sensors positions in a very short computational time and is insensitive to the non-submodularity of VoI.

Efficient Sensor Node Selection for Observability Gramian Optimization

“…. , n} for Equation (16), which requires computational complexities of O nr 3 for solving the algebraic Lyapunov equation (18) and obtaining the matrix multiplication inside, respectively. The first term in Equation ( 16) requires O n 2 r 2 ; therefore, the leading terms will be the sum O n 3 + O n 2 r 2 + O nr 3 per iteration.…”

“…Collecting information from sensor measurements is often the only viable approach when estimating the internal state or hidden physical quantities. The optimization of sensor positions was intensively discussed in order to determine the most representative sensors and to reduce the resulting estimation error, such as when monitoring sensor networks [1][2][3][4], fluid flows around objects [5][6][7][8][9][10][11][12][13][14][15], plants and factories [16][17][18], infrastructures [19][20][21], circuits [22], and biological systems [23], estimating physical field [24][25][26][27], and localizing sources [28,29]. Recent advances in data science techniques have enabled us to extract reduced-order models from vastly large-scale measurements of complex phenomena [30][31][32][33][34][35][36][37][38][39].…”