Real-time stochastic qualitative simulation of large scale air conditioning systems

Abstract

This paper gives a new algorithm for the efficient stochastic qualitative simulation, which can suppress the number of generated behaviors using the existing probabilities of states. In the new algorithm, the minimum existence probability, which is the lower bound of existence probability of states being preserved, is estimated with the Monte Carlo method. As a result, the states that are insignificant for envisioning useful behaviors are removed before generating them. In addition, the prefetch of measurement to check consistency of the states with the measurement in advance contributes to efficiency of the states generation. Through the simulation of the large scale building air conditioning systems, we clarify that the number of states can be reduced under computable order and the simulation terminates in real time. I IntroductionSince the system structure of building air conditioning systems is various and changeable, rule-based fault diagnosis is impractical because of frequent changes of rules. Modelbased approach[l, 21 is more suitable to diagnosis of such targets. In the model-based approach, a model for describing structures and functions of the target system is prepared in advance, and possible behaviors of the model of the system are derived. By comparing t h e behaviors derived based on the model with the observed behavior of IEEE Catalog Number: 95TH8081 the system, a part in failure is identified.Qualitative simulation is one of the effective methodologies to derive behaviors using model of the systerriI3, 4, 51. It has a n advantage of simulating the model without C O I I I~ plicated functions, because the elaborate physical mechanisms can be expressed with simple causal relations by using qualitative models.However, because of ambiguity caused by simplicity of the qualitative model, enormous number of behaviors a,re generated [6]. Especially, since the number of sensors in a. building air conditioning system is limited by constraint on costs, it is difficult t o prune a lot of improbable behaviors using real measured values.We have already proposed a stochastic qualita.tive simulation to cope with this problem[7, 81. In this method, all states, whose series specify the behaviors of'thc? system, have existence probabilities. Since states with relatively small existence probability are eliminated, explosive increase of behaviors can be prevented. Although this method has succeeded in reducing possible behaviors, however, the number of tenta.tive sta.tes generated before pruning becomes intractable in terms of memory c:;i.ixic:ity and execution time to apply the nietliod to largt? s(:aIc a i r conditioning systems.In this paper, we present a new algorithm for tlrt! stochastic q uali ta.t ive si r n 11 1a.t ion t 11 at ai iris ;t t niorc cfficient execution toward real-timc diagnosis of large: s(:i~Iv building air conditioning systcinis. This mcithotl introduces miriirnurri existencc proba.t>ility, which is tlcfin(id as

Application of non-intrusive polynomial chaos theory to real time simulation

“…For all the aforementioned reasons, real time simulation combined with statistical methods, can significantly increase coverage and statistical confidence, [4]- [6]. Sampling-based methods are adopted in most of situations.…”