Investigation of practical issues in building thermal parameter estimation

Large amounts of bacterial biomass are needed for the production of various bacterial products, especially for the production of recombinant proteins from E. coli-carrying recombinant DNA. The method of choice for obtaining the proper amount of the biomass is to grow the bacteria to high density, thereby increasing the production of bacterial biomass per unit of fermentor volume. The high density approach is of special importance when the recombinant proteins are produced after an induction step because the bacteria cannot be reused after the induction. There are several methods for obtaining high density growth of E. coli,14 all of which are based on supplying the growing bacterial culture with its nutrient needs. A nutrient of special importance is oxygen, which has limited solubility in aqueous solutions. Therefore, most of the high density methods specifically address the question of how to supply the needed soluble oxygen. The majority of the high density growth methods are based on a fed-batch approach that controls the carbon source (usually glucose) concentration of the media and thus also the bacterial growth rate and the oxygen c o n s ~ m p t i o n . ~ ~ ~ The main reason for using the fed-batch approach is to prevent inefficient utilization of the carbon source. The fed-batch control strategy is

Adaptive Control Strategy for Maintaining Dissolved Oxygen Concentration in High Density Growth of Recombinant E. coli

A heating system integrated into the building envelope, such as a floor radiant heating system, is defined as a building envelope heating system (BEHS). Thermal mass usually integrated with such a heating system can be utilized to lower both peak loads and operating costs and to reduce room temperature swings by predictive control while utilizing solar gains to reduce energy consumption. Nevertheless, techniques required for the predictive control of BEHSs need to be developed in order to materialize these potential benefits. A methodology proposed in this study integrates building thermal analysis and predictive control of BEHS. A computer method for generating the symbolic transfer function of buildings is developed as the first part of the methodology. It includes hybrid signal flowgraph and generalized-nodal admittance formulations, and an algebraic algorithm associated with the constraint conditions of inequalities. New concepts for thermal network modelling are presented, with which a combined thermal parameter such as the operative temperature can be explicitly represented with an imaginary thermal network. A building thermal system is systematically modelled with a generalized thermal network. Because some design parameters of interest, such as amount of thermal mass, can be kept as symbols in the model, sensitivity analysis, optimum design and control studies of building systems can be significantly facilitated. An optimal predictive control system developed integrates a weather predictor, set-point optimizer, a system identifier and an adaptive Generalized Predictive Control algorithm so as to achieve high building thermal performance. A new weather predictor, simplified through normalization, makes it feasible to quantify the qualitative weather forecast for solar radiation. Several implementation issues in on-line parameter estimation are investigated through experiments in an outdoor passive solar test-room. A Generalized Predictive Controller (GPC) with a feedforward control scheme is improved with a new algorithm. The zone set-point is optimized through the combination of dynamic programming and on-line simulation. The methodology has been verified with both experiments and simulations. Results show that the weather predictor is capable of generating reasonably accurate solar radiation and outdoor temperature profiles for one day. A building thermal model can be robustly identified under the supervision rules. The performance of GPC is superior to conventional on-off and PI controllers. The optimal set-point can be efficiently generated by the proposed approach, which may lead to large savings in operating energy costs when a BEHS is properly designed and operated.

A methodology for thermal analysis and predictive control of building envelope heating systems

The dissolved oxygen (DO) is an important variable in aerobic fermentations and affects the cell growth and product formation. Dissolved oxygen control is difficult in batch fermentations because of the time-varying conditions, time delays, and the probe dynamics. Modeling of the various patterns of biological activity in fermentations and their impact on the DO process dynamics is essential to both achieve a satisfactory control and to track the aforementioned patterns. An adaptive pole placement algorithm with time-delay compensation was used for controlling the DO, coupled with system identification using recursively estimated autoregressive models with exogeneous inputs (ARX). The flow rate of O2 in a constant flow rate gas inlet mixture is used as the manipulated variable. Supervision and coordination techniques are applied to improve the control performance. The control performance is affected by the accuracy of the model prediction and the selected time delay. The effect of DO level on the productivity of β-lactamase using Bacillus subtilis under oxygen-limited conditions is investigated. Beta-lactamase stability is improved under prolonged growth conditions with low DO levels. © 1998 John Wiley & Sons, Inc. Biotechnol Bioeng 60: 1–9, 1988.

Adaptive pole placement control algorithm for DO-control in β-lactamase production

The existence of a filter in the error path as in the case of active noise control has been shown to generally degrade the performance of the LMS algorithm. Thus the convergence rate is lowered, the residual power is increased and the algorithm can even become unstable. In order to stabilise the algorithm the input u(n) is filtered by a estimation of the error path filter leading to the FXLMS-algorithm. Modification of the FXLMS-algorithm assuring the well known properties of the LMS-algorithm by correcting the filtered error have also been reported. This correction of the filtered error also makes it possible to use RLS type algorithms for active noise cancellation. Common to all of the above cited algorithms is the need of a good estimate of the error-path filter. In this paper some possible alternatives that can do with very little knowledge of the error-path filter are analysed using a state space approach. >

Algorithms for active noise cancellation without exact knowledge of the error-path filter

Adaptive controllers have proven to work well as long as all theoretical pre-conditions are fulfilled and the free design parameters are chosen property. However in practice these pre-conditions may be violated and/or some free design parameters are not chosen properly. Both result in a lowering of the control performance. Therefore a supervision and coordination level is introduced, monitoring the adaptive controller and taking appropriate actions if malfunctions are detected. The introduced supervision and coordination level exists altogether of two parts, named start-up part and run-time part. The start-up part involves tasks for the configuration and the implementation of adaptive controllers and the run-time part involves tasks for the supervision and coordination of the adaptive controller in closed-loop operation. A sensible start-up procedure is recommended and an eigenvalue analysis is shown in order to monitor the parameter estimation. To track time-varying process parameters an eigenvalue controlled forgettig factor. is introduced. Simulation and experimental results for an a.c. motor are shown, indicating an improvement of the overall control performance.

Supervision and Coordination of Parameter-Adaptive Controllers

Knowledge Based Adaptive Control with Learning and Intelligent Abilities

Intelligence and learning are keywords which are more and more common in the area of adaptive control. From a control engineering point of view, here the meaning of intelligence and learning could be characterized by automatically finding both an accurate process model and an optimal control algorithm and by features like fault detection and supervision of the process and its control. This also includes start-up of an adaptive controller. In this paper different methods will be presented to deal with simple and complex processes. A knowledge based adaptive controller will be discussed which leads the control engineer to the optimal control algorithm, independent of the type of process. Several simulation and experimental results will be shown for critical processes.

T. Knapp

Papers

Supervision and Coordination of Parameter-Adaptive Controllers

Knowledge Based Adaptive Control with Learning and Intelligent Abilities

Supervision and Coordination of Parameter-Adaptive Controllers

Knowledge based adaptive control with learning and intelligent abilities