Feedforward Feedback circuits are superior

  • Everyone is familiar with feedback circuits for control. PID is an example of a feedback circuit. It is mainly control theory specialists who are acquainted with the notion that feedback combined with feedforward makes for superior control.


    Feedback control simply uses the error signal from the controlled variables to create a command for the actuator. A feedforward circuit measures the disturbance to the system to guess at the command for the actuator. By combining the feedback error signal with the feedforward guess, far superior control can be achieved. In the instance of a LENR system, the disturbance could be radioactive signal from the energy generating mechanism. By using that "disturbance" the actuator (e.g. the external heater) can be commanded BEFORE the temperature is even conducted to the temperature measuring device.


    Much literature is available on combined feedforward / feedback control. Here is a starter link: https://en.wikipedia.org/wiki/Feed_forward_(control)


    I invite comment from control theory specialists for the special case of LENR.

  • To make a good predictor, you need to have a good model of the system you control, or a good adaptative model.
    For LENR which is very non-linear, is not easy, but it will be very inspiring for scientists.
    It can also behave much better than linear feedback because of such non linearity.

  • Yes AlainCo . Here is an example of non-linear control for a LENR reactor: if a burst of radioactivity is detected, a comparator switches, the comparator triggers a monostable that shuts off (or reduces) the control actuator (i.e. external heat, or, nuclear control rod , RF stimulus etc.) . The monostable should probably be set to a time about equal to or slightly exceeding the heat conduction time to the control thermocouple sensor.

  • You raise an interesting idea for people trying to control a reactor.
    like when you control a heater for a house, you need many sensors, and to learn the model of your "problem".
    one things to do is to try to have sensors that react very quickly, beside those that are near the real goal.
    For example for heater you have sensor in the boiler, and some in the rooms, and even some outside.
    For servomotors you have often a control loop on current (torque), and one on position that embed the torque loop.
    It maybe sensors in the reactor, near the source, or as you propose sensors detecting high-energy radiation, why not optical responses.
    the idea is that systems are more easy to control if the sensors are reacting if not linearly, at least continuously/monotonously to the goal.
    For example I would advise people to measure low energy gamma, UV, light, IR, electromagnetic energy, and to see if those fast response sensors give a better vision on the phase-space of the system.


    What you seems to present is more a safety system, and the response may not be a shutdown (may make things worse or just prevent any usage), but a "behavior that works" based on experiments (like heating for, faster, slower, changing rhythm...).


    much work to learn how the reactor response.
    This is more about zoology, and my 2cent advice is to let physicist only advise how to install sensors, measure quantities, and to avoid health risk, but not give any theory.
    Control-command is a behaviorist territory.


    my 4c cent bet is that if someone succeed in describing the LENR animal like a zoologist (ethologist?), he may give huge data to the physicists. the behavior of the system seen without any preconception, can be a key information for theorist.