Uncertainty and optimization in the design of building subsystems

  • Authors:
  • Elliott E. Dudnik

  • Affiliations:
  • -

  • Venue:
  • DAC '77 Proceedings of the 14th Design Automation Conference
  • Year:
  • 1977

Quantified Score

Hi-index 0.00

Visualization

Abstract

Problems of building system design often are based upon not only a complete knowledge of the physical system and its capacities but also estimates or perceptions of the user requirements and external conditions. While the process of design implies a fit between the user demands or requirements and system conditions and constraints, the lack of accurate information about these user requirements and demands alters the quality of the final product. Since we generally neglect to consider that many of the design problem parameters are not completely known or deterministic, but are, in fact, estimated or stochastic, we often fail to recognize that our final design may not have completely met the actual user needs or demands. Failure to have achieved the user needs may result in systems that either cannot perform adequately under severe operating conditions or which may cause physical damage or harm. Therefore, most design procedures and design codes provide an adequate level of safety to ensure against personal harm or injury and to minimize risk and property damage, they do so, however, by over-designing the components of the system rather than by dealing with the actual problem of the uncertainty which underlies the problem parameters. Thus, while the design methods themselves have become increasingly sophisticated, the design parameters are often still estimates and do not truly warrant the treatment which they receive in subsequent calculations. The harm in such an approach is not so much in the use of safety factors or other such devices to accommodate the stochastic nature of the problem, but in the fact that the final design is usually overdesigned and results in a waste or misuse of equipment, material, energy, space or money. Given the increased awareness of the need to conserve material or energy resources, client-owner-users are no longer willing to pay the “premium” that results from providing such system overcapacity or over design. With no direct benefit resulting from such an additional expenditure of money or resources, the demand and need for greater efficiency and economy is necessary while still acknowledging, of course the need for safety and precaution. What is necessary, therefore is a design procedure which provides the means for system design that will allocate or distribute the available resources in a manner which will minimize costs and/or maximize user gains while recognizing that the demands upon the system are based upon estimates and thus are subject to varying degrees of risk or uncertainty.