Font Size: a A A

Modeling And Optimal Control Mixed With Logic And Rules In Industrial Process

Posted on:2005-01-27Degree:DoctorType:Dissertation
Country:ChinaCandidate:L Y ZhangFull Text:PDF
GTID:1118360122987919Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Hybrid systems method is the result of discrete event systems research and need of modern industrial process control. Exploiting hybrid systems method, industrial process control can take full advantage of artificial intelligence and computational intelligence, operational research and application logic. Then we can advance new concept of solving the problem of modeling and optimal control in large systems. This paper build industrial process hybrid model mixed with logical and rules and advance mixed integer nonlinear programming algorithms combined with constraint programming. At last the author apply those methods to industrial process systems.The main contribution and research works are as follows:1.Based on mechanism modeling basic state space equation was built. And then by exploiting expert knowledge, operational experience and logical constraints in systems, industrial process hybrid model was built. Stability of this model was defined and the method of designing controller based on predictive control was present. At last the stability condition of this predictive control method was gived.2.This paper advanced mixed integer nonlinear programming (MINLP) algorithms mixed with constraint programming (CP) to solve industrial process hybrid model mixed with logical rule. CP methods have proved to be successful in solving highly constrained discrete optimization and feasibility problems. It can effectively solve hybrid model mixed with logical rule by integrating CP, nonlinear programming and mixed integer linear programming. Computational results demonstrated this algorithm.3.Apply industrial process hybrid model mixed with logical and rule to simplify activated sludge process model in wastewater treatment and deduce to a reduced model adapted to control. The simulation result show that this method can combine mechanism model of activated sludge process and expert knowledge and operational experience and can get better effect of optimization and control.Moreover,the computational time is short and suitable for real-time control in wastewater treatment by applying mixed integer nonlinear programming mixed with logical rules.4.Interval Temporal Logic (ITL) is a temporal logic which includes a basic construct for the sequential composition of two formulas as well as an analog of Kleene star. In this paper,Interval Temporal Logic is applied to represent the hybrid systems. This method is demonstrated by an example in water industry systems.5.Optimization of N-removal in sewage treatment is accomplished by exploiting mixed logical dynamic method (MLD). According to simplifying activated sludge process No.1 model (ASM1),N-removal dynamic models in continuously mixed,continuously fed activated sludge process (ASP) are established by employing expert experience about activated sludge process. Then predictive control approach is applied to optimal control of this process. The simulation results show that it can cover the relative expert experiences more widely by applying MLD to modeling and control of activated sludge process. These expert experiences combined with continuous variable model make model more precise and can get better effect of optimization and control. This study provides a new approach to the research of sewage treatment.
Keywords/Search Tags:hybrid system, constraint programming, mixed integer nonlinear programming, predictive control, logical and rules
PDF Full Text Request
Related items