Abstract: An adaptive real time thermal processing system is presented that includes a multivariable controller. The method includes creating a dynamic model of the MLD processing system and incorporating virtual sensors in the dynamic model. The method includes using process recipes comprising intelligent set points, dynamic models, and/or virtual sensors.

Abstract: An adaptive real time thermal processing system is presented that includes a multivariable controller. The method includes creating a dynamic model of the MLD processing system and incorporating virtual sensors in the dynamic model. The method includes using process recipes comprising intelligent set points, dynamic models, and/or virtual sensors.

Abstract: An adaptive real time thermal processing system is presented that includes a multivariable controller. Generally, the method includes creating a dynamic model of the thermal processing system; incorporating reticle/mask curvature in the dynamic model; coupling a diffusion-amplification model into the dynamic thermal model; creating a multivariable controller; parameterizing the nominal setpoints into a vector of intelligent setpoints; creating a process sensitivity matrix; creating intelligent setpoints using an efficient optimization method and process data; and establishing recipes that select appropriate models and setpoints during run-time.

Abstract: An adaptive real time thermal processing system is presented that includes a multivariable controller. Generally, the method includes creating a dynamic model of the thermal processing system; incorporating reticle/mask curvature in the dynamic model; coupling a diffusion-amplification model into the dynamic thermal model; creating a multivariable controller; parameterizing the nominal setpoints into a vector of intelligent setpoints; creating a process sensitivity matrix; creating intelligent setpoints using an efficient optimization method and process data; and establishing recipes that select appropriate models and setpoints during run-time.

Abstract: Methods for adaptive real time control of a system for thermal processing substrates, such as semiconductor wafers and display panels. Generally, the method includes creating a dynamic model of the thermal processing system, incorporating wafer bow in the dynamic model, coupling a diffusion-amplification model into the dynamic thermal model, creating a multivariable controller, parameterizing the nominal setpoints, creating a process sensitivity matrix, creating intelligent setpoints using an efficient optimization method and process data, and establishing recipes that select appropriate models and setpoints during run-time.

Abstract: A heat treatment apparatus for making a heat treatment while estimating temperatures of objects-to-be-processed that can estimate correct temperatures of the objects-to-be-processed. A reaction tube includes heaters and temperature sensors, and receives a wafer boat. A controller estimates temperatures of wafers and temperatures of the temperature sensors in zones in the reaction tube corresponding to the heaters by using the temperature sensors and electric powers of the heaters. Based on relationships between estimated temperatures of the temperature sensors and really metered temperatures, functions expressing the relationships between the estimated temperatures and the really metered temperatures are given for the respective zones. The functions are substituted by the estimated wafer temperatures to correct the estimated wafer temperatures.

Abstract: There is provided a batch type heat treatment system, control method and heat treatment method capable of appropriately coping with a multi-product small-lot production. A reaction tube 2 comprises a plurality of heaters 31 through 35 and a plurality of temperature sensors, and houses therein a wafer boat 23. A control part 100 stores therein many mathematical models for estimating (calculating) the temperature of wafers W in the reaction tube 2, in accordance with the number and arranged position of the wafers W mounted on the wafer boat 23, and many target temperature trajectories. If the wafer boat 23 is loaded in the reaction tube 2, a mathematical model and a target temperature trajectory corresponding to the number and arranged position of the mounted wafers W are read.

Abstract: The present invention includes a method and system for reducing switching transients via large signal/small signal thresholding. The present invention prevents a switch from occurring if when the system (i.e., the system under control) is in small signal mode and permits switches to occur when the system is in large signal mode. Large signal and small signal modes are determined by comparing the difference between the most current value of a comparator variable and the final setpoint with a threshold distance. The system is in small signal mode if the most current value of the comparator variable less the final setpoint value is less than the threshold distance. Otherwise, the system is in large signal mode.

Abstract: A method for computer-aided design. The method includes the steps of representing a computer-aided design activity as a design cycle and defining a goal of the design activity. The method also includes the steps of defining alternate design steps toward achieving the goal. According to one embodiment, the design activity is represented as a graph containing the alternate design steps. According to this embodiment, a cost metric is assigned for each of the alternate design steps that includes elements representing relative costs of taking each of a plurality of paths in the graph. Element values and bounds on the cost metric are the determined by searching the graph to project costs incurred upon execution of each of the alternate design steps to achieve the goal. The cost of achieving the goal is then optimized by selecting one of the alternate design steps associated with a smallest cost metric.

Abstract: The present invention is a method and apparatus for dynamic optimization of a dynamic physical system having a device under control and a controller along with a modeling system having an estimator and a constrained nonlinear program. The time varying trajectories of the dynamic physical system are parameterized and run through the modeling system in order to produce a set of optimal time varying setpoint trajectories to be used in the dynamic physical system. The optimal time varying setpoint trajectories optimize the physical system to produce the desired end result (or product) of the dynamic physical system.

Abstract: A method for the estimation of the state variables of nonlinear systems with exogenous inputs is based on improved extended Kalman filtering (EKF) type techniques. The method uses a discrete-time model, based on a set of nonlinear differential equations describing the system, that is linearized about the current operating point. The time update for the state estimates is performed using integration methods. Integration, which is accomplished through the use of matrix exponential techniques, avoids the inaccuracies of approximate numerical integration techniques. The updated state estimates and corresponding covariance estimates use a common time-varying system model for ensuring stability of both estimates. Other improvements include the use of QR factorization for both time and measurement updating of square-root covariance and Kalman gain matrices and the use of simulated annealing for ensuring that globally optimal estimates are produced.

Abstract: A method for computer aided design of a product or process. The method includes the steps of representing a computer-aided design activity as a design cycle. The design cycle is then converted into individual iterations of decisions to be made within the design cycle. The individual iterations of decisions are then mapped into a hierarchical structure of influence diagrams. The decision making process of the influence diagrams are performed and data is collected from the decision process that has been performed. The remaining unexecuted decisions and their posterior probabilities are updated based upon the outcomes of the decision process that was most recently performed.

Abstract: The present invention is a method and apparatus for dynamic optimization of a dynamic physical system having a device under control and a controller along with a modeling system having an estimator and a constrained nonlinear program. The time varying trajectories of the dynamic physical system are parameterized and run through the modeling system in order to produce a set of optimal time varying setpoint trajectories to be used in the dynamic physical system. The optimal time varying setpoint trajectories optimize the physical system to produce the desired end result (or product) of the dynamic physical system.

Abstract: A method for processing signals and controlling a physical system in which measurements are obtained at different time scales and/or different space scales. Signals generated from the physical system are processed by first creating a first dynamic model at a first time/space scale that consists of a first set of parameters, a first set of states and a first set of inputs. A second dynamic model at a second time/space scale is also created and consists of a second set of parameters, a second set of states and a second set of inputs. At least one of the first set of parameters in the first dynamic model are computed from the second set of states in the second dynamic model. A second estimator is then created to produce estimates of the second set of states using the second dynamic model, measurements from the first physical system at the second time scale.

Abstract: A method for the estimation of the state variables of nonlinear systems with exogenous inputs is based on improved extended Kalman filtering (EKF) type techniques. The method uses a discrete-time model, based on a set of nonlinear differential equations describing the system, that is linearized about the current operating point. The time update for the state estimates is performed using integration methods. Integration, which is accomplished through the use of matrix exponential techniques, avoids the inaccuracies of approximate numerical integration techniques. The updated state estimates and corresponding covariance estimates use a common time-varying system model for ensuring stability of both estimates. Other improvements include the use of QR factorization for both time and measurement updating of square-root covariance and Kalman gain matrices and the use of simulated annealing for ensuring that globally optimal estimates are produced.

Abstract: A method for computer-aided design. The method includes the steps of representing a computer-aided design activity as a design cycle and defining a goal of the design activity. The method also includes the steps of defining alternate design steps toward achieving the goal. According to one embodiment, the design activity is represented as a graph containing the alternate design steps. According to this embodiment, a cost metric is assigned for each of the alternate design steps that includes elements representing relative costs of taking each of a plurality of paths in the graph. Element values and bounds on the cost metric are the determined by searching the graph to project costs incurred upon execution of each of the alternate design steps to achieve the goal. The cost of achieving the goal is then optimized by selecting one of the alternate design steps associated with a smallest cost metric.

Abstract: A system for controlling a thermal reactor is disclosed that characterizes the thermal reactor with a reactor model that indicates behavior of the thermal reactor and of a load contained in the thermal reactor and that accounts for interaction among a set of heating zones of the thermal reactor. An online reactor model is then determined that estimates the thermal behavior of the load based upon an online input power to the thermal reactor and upon an online temperature indication from the thermal reactor. A time varying temperature and reactant flow recipe is determined that minimizes end of run parameters on the load. A multi-variable controller is employed to minimize temperature deviations of the load from a predetermined temperature recipe or time varying trajectory.