Search

The purpose of automatic controls is to modulate the capacity of HVAC equipment to satisfy the heating and cooling requirements of a building and to monitor and control safely the operation of HVAC systems. This type of control traditionally has been performed by pneumatic and electric control systems. Today, microprocessors are taking the place of conventional controllers and are rapidly becoming the preferred automatic control method for building HVAC systems. Because of the declining cost of semiconductor chips and the proven reliability of microprocessor-based controllers, direct digital control (DDC) systems now compare favorably against conventional control methods in both functional and economic terms. DDC, however, is not the answer to every automatic control problem. A solid grounding in control system fundamentals is necessary before one can fully understand how to best apply direct digital controllers to common control system designs. Because DDC systems are typically used in combination with other types of control systems, control system fundamentals are now reviewed.

Regardless of whether you are a consultant, a contractor, or a building owner attempting to upgrade or replace an outmoded control system, it is absolutely necessary to develop a specification to describe the system you want and how it should perform. There are few standard specifications available for computer-based control systems. Many that are available from industry and professional sources have been developed from negative past experiences with products and their installing contractors and are very conservative and quite defensive.

The computer is the heart of any direct digital control (DDC) system. Although many models of DDC computer systems are currently available, most operate in the same basic way and are similar in design to personal computers (PCs). At the most basic level, all computers consist of hardware and software.

Direct digital control (DDC) is a control process in which a microprocessor controller constantly updates an internal information database by monitoring information from a controlled environment and continuously produces corrective output commands in response to changing control conditions. This process is similar to the way traditional control systems work, except that digital information processing produces a much faster and more accurate control system response under all conditions at all times.

Electrical services are becoming more and more important in modern intelligent buildings as most of the conventional mechanical systems, i.e. pneumatic and hydraulic, have already undergone technological revolutions, being upgraded to the direct digital control mode. In this chapter, we shall look at selected applications of AI techniques in electrical systems, including computer vision based metering and lighting control, fuzzy logic based harmonics patterns recognition, neural network based harmonics evaluation and lighting design using computer algebra.

The continuing evolution of microprocessor design and manufacture has fostered a similar evolution in the development of direct digital controllers. As large-scale integration and lower cost have made microprocessor power accessible and affordable to manufacturers of all kinds of devices, the usage of microprocessor intelligence has been driven down to the lowest possible levels of modern control systems. In short, intelligence is everywhere. What remains lacking in the digital control systems of today is the ability to connect all of this intelligence in a way that is meaningful and valuable to building operators.

In Chapter 7, a methodology for designing direct digital control (DDC) systems was suggested to help organize the control design process. The point was made that design decisions are affected by economics– the first cost of the system versus the benefits that it provides for the purpose of reaching a balance between function and affordability.

The topic of Direct Digital Control (DDC) is described here as a distinct concept. It is important to distinguish between the Energy Management and Control System (EMCS) which has been described as a computer, and DDC which is a method of processing and control. DDC control strategies are employed by many EMC systems today and they serve to enhance the control sophistication of these systems.

The structure of controllers for the direct digital control of power in a nuclear reactor is examined and features possibly important to licensing are identified. This was done by first reviewing the human approach to process control and then discussing designs whereby digital systems could be made to emulate each of the functions that a human operator now performs. Features identified as relevant to licensing include the separation of the safety and control systems, the use of supervisory algorithms that incorporate the concept of ‘feasibility of control’, signal validation, verification of signal implementation, and the use of either fault-tolerant or redundant hardware. Examples supporting the above findings are drawn from the MIT experience in first designing and then licensing a digital, closed-loop controller for the 5 MWt MIT Research Reactor.

A Level III Automation or EMC system is also referred to in this book as an engineering control center. This is because these systems perform a central engineering function, similar to the concept of a Level II system, along with automatic control, energy monitoring, maintenance scheduling, information management, life safety and other functions which are made possible by a software-based logic system. The current state of the art in this technology is the Direct Digital Control (DDC) system. DDC is treated separately in Chapter 10. A further enhancement is distributed processing which will be discussed later in this chapter.

The advent of the minicomputer is having a profound influence on real-time applications such as data acquisition, direct digital control, supervisory control, manufacturing monitoring and control, message switching, and communication systems. It is not simply the low cost of the minicomputer itself that makes it attractive, but rather the modular solution to real-time systems problems it provides. That is, the major difference between a large computer and a minicomputer is not the power of the instruction sets. It is rather dedicated use of the machine, which in turn leads to software engineered specifically for the application.