Design of an AI-Based self-sustaining habitats control system

The work described here derives from the implementation of intelligent control for a Bioregenerative Closed Ecological Life Support System. The AI-Based, Distributed Environment Control System (A1DECS) will be capable of supporting, in an integrated fashion, all activities ranging from planning of agricultural activities at the highest level to real-time control of environmental conditions at the lowest level. This paper describes the AIDECS subsystem that deals with control over two widely different time scales: 1) that of scheduling crop planting and harvesting over the extended horizon needed to ensure maintenance of the CO2/O2 and other gas balances (typically involving units of weeks, months and years), and 2) that of controlling environmental parameters such as temperature and humidity to be properly correlated with crop requirements and external weather conditions (typically involving units of hours and days). INTRODUCTION "Intelligent control," the intersection of artificial intelligence (AI), conventional automatic control, and operations research approaches, is receiving increasing attention in both theory and application[I]. This paper describes the AI-Based, Distributed Environment Control System (AIDECS) under development for a Bioregenerative, closed ecological life support system (BCELSS) at the Environmental Research Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery. To copy otherwise, or to republish, requires a fee and/or specfic permission. © A C M 1 9 8 8 0 8 9 7 9 1 2 7 1 3 / 8 8 / 0 0 0 6 / 1 0 5 9 $1.50 Lab of the University of Arizona. The AIDECS will be capable of supporting, in an integrated fashion, all activities ranging from planning of agricultural activities at the highest level to real-time control of environmental conditions at the lowest level. This paper describes the AIDECS subsystem that deals with control over two widely different time scales: 1) that of scheduling crop planting and harvesting over the extended horizon needed to ensure maintenance of the CO2/O2 and other gas balances (typically involving units of weeks, months and years), and 2) that of controlling environmental parameters such as temperature and humidity to be properly correlated with crop requirements and external weather conditions (typically involving units of hours and days). The AIDECS subsystem contains a natural languagelike interface in which the wide variety of schedules required in the BCELSS may be specified in a uniform manner. Such specifications are mapped into schedule objects that are stored for later implementation and reuse. Such objects may include specification of events which put into effect other schedule objects. This gives rise to hierarchical schedules which facilitate timing of events to occur at coarseand fine-grained time units. Likewise, several schedule objects may be put into effect at the same time, thus enabling scheduling of concurrent, correlated activities. A hierarchical schedule is implemented by first interpreting its root schedule object. Such an interpretation is performed by mapping the schedule into a set of elementary, rule-like "activities" to be evaluated by an evaluator, in similar fashion to that of a conventional inference engine. The evaluator continually cycles through its list of activities, checking whether the timing and preconditions of activities are satisfied, and executes the actions of those that are so triggered. Actions at the