Individual differences |
Methods | Statistics | Clinical | Educational | Industrial | Professional items | World psychology |
Feed-forward is a term describing a kind of system which reacts to changes in its environment, usually to maintain some desired state of the system. A system which exhibits feed-forward behavior responds to a measured disturbance in a pre-defined way — contrast with a feedback system.
Many prerequisites are needed to implement a feed-forward control scheme: the disturbance must be measurable, the effect of the disturbance to the output of the system must be known and the time it takes for the disturbance to affect the output must be longer than the time it takes the feed-forward controller to affect the output. If these conditions are met, feed-forward can be tuned to be extremely effective.
Feed-forward control can respond more quickly to known and measurable kinds of disturbances, but cannot do much with novel disturbances. Feed-back control deals with any deviation from desired system behavior, but requires the system's measured variable (output) to react to the disturbance in order to notice the deviation.
Feed-back control is exemplified by homeostatic regulation of heartbeat in response to physical exertion. Feed-forward control can be likened to learned responses to known cues. These systems could be in control theory, physiology or computing.
Elements common to all feed-forward systemsEdit
A feed-forward system can be illustrated by comparing it with a familiar feedback system — that of cruise control in a car. When in use, the cruise control enables a car to maintain a steady road speed. When an uphill stretch of road is encountered, the car slows down below the set speed; this speed error causes the engine throttle to be opened further, bringing the car back to its original speed (a PI or PID controller would do this. Note that a good PID control will return the car to the original speed, after an initial transient response).
A feed-forward system on the other hand would in some way 'predict' the slowing down of the car. For example it could measure the slope of the road and, upon encountering a hill, would open up the throttle by a certain amount, anticipating the extra load. The car does not have to slow down at all for the correction to come into play.
However, other factors than the slope of the hill and the throttle setting influence the speed of the car: air temperature, pressure, fuel composition, wind speed, etc. Just setting the throttle based on a function of the slope may not result in constant speed being maintained. Since there is no comparison between the output variable, speed, and the input variable, it is not possible to resolve this problem with purely feed-forward control.
Fortunately, the two types of control are not mutually exclusive; the feed-forward system just described could be combined with the feed-back system of conventional cruise control to allow quick response with the feedback system cleaning up for any error in the predetermined adjustment made by the feed-forward system. See Model predictive control.
Feed-forward does not have the stability problems that feed-back can have. Feed-forward needs to be a pre-calibrated cause → effect, feed-back does not. This is another way of saying what was said above - that feed-forward control applies to measurable disturbances with known effects.
In Behavioural and cognitive scienceEdit
Feedforward, Behavior and Cognitive Science is a method of teaching and learning that illustrates or indicates a desired future behavior or path to a goal. Feedforward provides information, images, etc. exclusively about what one could do right in the future, often in contrast to what one has done in the past. The feedforward method of teaching and learning is in contrast to feedback concerning human behavior because it focuses on learning in the future, whereas feedback uses information from a past event to provide reflection and the basis for behaving and thinking differently. In isolation, feedback is the least effective form of instruction, according to US Department of Defense studies in the 1980s. Feedforward is the opposite of feedback, and was coined by Peter W. Dowrick in his dissertation.
One concept of feedforward originated in behavioral science. Related concepts have emerged in biology, cybernetics, and management sciences (see separate entries in Wikipedia). Since the 1970s, the understanding of feedforward has evolved to become more explicit, more useful, and to help the understanding of brain function and rapid learning. The concept contributed significantly to research and development of video self modeling (VSM). The most productive advances in feedforward came from its association with videos that showed adaptive behavior one had never exhibited before, at least not in the context shown in the video (see Dowrick, 1983, pp. 111, 121; 1991, pp. 110–3, 120-2, 240-1; 1999, esp. pp. 25–26). For example, a boy with autism role-plays squeezing a ball (stress management technique) instead of having a tantrum when his work is found imperfect by the teacher – or a selectively mute child is seen on video talking at school (which she never did), by editing in footage of her talking at home (location disguised by use of a classroom backdrop). By selectively editing a video, a clip was made that demonstrated the desired behavior and allowed the children to learn from their future successes.
By reference to its historical context of VSM, it became recognized that feedforward comprised component behaviors already in the repertoire, and that it could exist in forms other than videos. In fact, feedforward exists as images in the brain, and VSM is just one of many ways to create these simulations. The videos are very short – the best are 1 or 2 minutes long, and achieve changes in behavior very rapidly. Under the right conditions, a very few viewings of these videos can produce skill acquisitions or changes in performance that typically take months and have been resistant to change by other methods. The boy with autism and the girl with selective mutism, mentioned above, are good examples. Further examples can be found in journal articles, and on the web (e.g., in sport, http://keithlyons.wordpress.com/2009/06/28/feedforward/).
The evidence for ultra-rapid learning, built from component behaviors that are reconfigured to appear as new skills, indicates the feedforward self model mechanism existing in the brain to control our future behavior  See Figure 1. [insert Figure 1 about here]. That is, if the conditions of learning are right, the brain takes pieces of existing skills, puts them together in new ways or in a different context, to produce a future image and a future response. Thus we learn from the future – more rapidly than we learn from the past. Further evidence comes from cognitive processes dubbed “mental time travel”  and for parts of the hippocampus etc. where they occur. However, the links between these hot spots in the brain and feedforward learning have yet to be confirmed.
Feedforward concepts have now become firmly established in at least four areas of science, and they continue to spread. Feedforward often works in concert with feedback loops for guidance systems in cybernetics or self-control in biology (**insert link). Feedforward in management science enables the prediction and control of organizational behavior. These concepts have developed during and since the 1990s. Feedforward in procedures of behavior change and rapid learning have been quietly with us since the mid 1970s. In summary, feedforward in behavioral and cognitive science may be defined as “images of adaptive future behavior, hitherto not mastered”; images capable of triggering that behavior in a challenging context. Feedforward is created by restructuring current component behaviors into what appears to be a new skill or level of performance.
Physiological feed-forward systemEdit
Main article: Feed-forward (physiology)
In physiology, (also called a feed-forward homeostatic control system) is a homeostatic control system in which, the anticipatory effect that one intermediate exerts on another intermediate further along in the pathway allows the system to anticipate changes in a regulated variable.
Gene regulation and feed-forwardEdit
The cross regulation of genes can be represented by a graph, where genes are the nodes and one node is linked to another if the former is a transcription factor for the latter. A motif which predominantly appears in all known networks (E.Coli, Yeast,...) is A activates B, A and B activate C. This motif has been shown to be a feed forward loop, detecting non-temporary change of environnement.
Feed-forward systems in computingEdit
Main article: Perceptron
- S. Mangan A. Zaslaver & U. Alon, "The coherent feed-forward loop serves as a sign-sensitive delay element in transcription networks", J. Molecular Biology 334:197-204 (2003).
|This page uses Creative Commons Licensed content from Wikipedia (view authors).|
<ref>tags exist, but no
<references/>tag was found