A Pilot Study Of Scheduling, Tracking And Related Control Problems In The Presence Of External Environment Changes

In industrial production, the control target is to ensure the controlled output to track the preset operat-ing status, which is the so-called "set point". Take the continuous stirred tank reactor (CSTR) in chemical processes for example. The control objective is to guarantee that ingredients can reach the required tempera-ture and concentration to develop the qualified chemical products. In the past set-point control research, the set point (i.e., reference signal) is generally assumed to be known a priori. Very few researchers focus on how to determine the set point. In this dissertation, we attempt to study the scheduling, tracking and related control problems based on external environment changes, and execute a tentative exploration in the problem how to plan the operational trajectory of the controlled object according to dynamical changes of industrial production environment. The main content is composed of the following three respects:1.) In order to plan the operational trajectory of the production process based on external environmen-t changes, the first task is to investigate dynamic scheduling in the presence of external environment changes.(1) In real factories, human schedulers often perceive and deal with external environment changes, which present risks to the production processes or product quality. In the second chapter we pro-pose a new rescheduling framework to cope with external environment-related risks. The extended rescheduling environment is introduced to identify various types of rescheduling analysis triggers, which is composed of three elements:Case, Resource and Job. Case is designed to describe the ex-ternal environment. Then, anew rescheduling judgement measure-suitability is developed to judge whether to reschedule or not when the extended rescheduling environment will change. Finally, the machine shop scheduling problem of a wood furniture factory is provided to show the importance of considering external environment in handling production scheduling problems.(2) In the third chapter, we design a new dispatching rule for the job sequencing problem in the machine shop of a wood furniture factory. The main aim is to reduce the possibility that quality problems happen resulting from "bad" weather. First, two indices-risky duration and risk magnitude are established to characterize the weather conditions. Based on the two indices, the job suitability is derived by using the fuzzy decision method, and integrated with a traditional heuristic to compute the dispatching priority of each job. Then, we design a measure-matching degree to evaluate the effectiveness of our dispatching rule. Finally, large-scale experiments are provided to show that our dispatching rule can greatly increase the matching degree while maintaining low weighted tardiness.(3) Based on the previous chapter, we improve dispatching rules for the job sequencing problem in the machine shop of a wood furniture factory. The main aim is still to reduce the possibility that quality problems happen resulting from "bad" weather. Noting that the sensitivity of different jobs to weather is different, we introduce another new index-job sensitivity. According to all the three indices, the job suitability is derived by using the fuzzy decision method, and integrated with tra-ditional heuristics to compute the dispatching priority of each job. And a suitability tunning factor is established to regulate the magnitude of the job suitability. Then, we design a measure-moisture penalty to evaluate the effectiveness of our dispatching rules. Finally, large-scale experiments are provided to show that our dispatching rules can greatly reduce the risk that quality problems occur while maintaining low weighted flowtime/tardiness.2.) Because external environment changes, some properties of the control object have changed such that the system performance cannot be realized. As a result, the control target cannot be completed. An important task is to find what references the controlled object can reach. In the fifth chapter, the output tracking control problem is investigated for a class of nonlinear discrete-time systems via T-S fuzzy observer-based control approach. A delayed fuzzy observer is firstly established to estimate state variables and premise variables. Then, a target state is built to convert the output tracking problem into the stabilization problem of a tracking error system. As a result, we can use the approaches for stabilization problems to realize the output tracking. Finally, a new delayed nonquadratic Lyapunov function is introduced to design controller gains and observer gains separately. A practical example, mass-spring-damper system, is considered to show the effectiveness of our results. A numerical example demonstrates the superiority of the proposed method.3.) In the end, the stability and stabilization problems are considered for discrete-time switched systems in both nonlinear and linear contexts. By introducing the concept of forward average dwell time and using multiple-sample Lyapunov-like functions variation, the extended stability results for discrete-time switched systems in the nonlinear setting are first derived. Then, the criteria for stability and stabilization of linear switched systems are obtained via a newly constructed switching strategy, which allows us to obtain exponential stability results. A new kind of mode-dependent controller is designed to realize the stability conditions expressed by the multiple-sample Lyapunov-like functions variation. Based on this controller, the stabilization conditions can be given in terms of linear matrix inequalities (LMIs), which are easy to be checked by using recently developed algorithms in solving LMIs. Finally, a numerical example is provided to show the validity and potential of the results.