Combined proportional plus integral (PI) and neural network (nN) controller

A neural network controller in parallel with a proportional-plus-integral feedback controller in a control system, the system comprising:

• a plurality of input ports of the neural network controller, each said input pod for receiving a respective input signal representing a respective condition of a process;
• a first set of data comprising a plurality of learned output values of the neural network controller obtained during a training period thereof using a plurality of first inputs representing a plurality of conditions of said process;
• and
• in operation, the neural network controller to contribute to a current output, Oτ, of the proportional-plus-integral feedback controller upon detection of at least one triggering event, said event comprising a change in any one of said respective input signals at a time, τ, greater-than a preselected amount indicating a condition deviation;
• wherein said operation of the neural network controller in communication with the proportional-plus-integral feedback controller, further comprises:
• (a) upon said detection at said time, τ, a respective one of said learned output values, ONN, of said first set of data corresponding with said condition deviation is added to the proportional-plus-integral feedback controller current output Oτ, and at said time, τ, a prior control output value, Oτ−1, of the proportional-plus-integral feedback controller associated with a prior time, τ−1, does not contribute to said current output, Oτ;
• (b) otherwise, the system operates with the neural network controller making no contribution to the proportional-plus-integral feedback controller current output Oτ;
• and
• (c) the proportional-plus-integral feedback controller functioning with a proportional gain constant Kp and an integral gain constant Ki that remain unchanged during said operation.

A method for controlling a process with a neural network controller operating in parallel with a proportional-plus-integral feedback controller, the method comprising the steps of:

• generating a first set of data comprising a plurality of learned output values of the neural network controller obtained during a training period thereof using a plurality of first inputs representing a plurality of conditions of a process;
• receiving, at each of a plurality of input ports of the neural network controller, an input signal representing a respective condition of said process;
• and
• the neural network controller operating in parallel with the proportional-plus-integral feedback controller to contribute to a current output, Oτ, of the proportional-plus-integral feedback controller upon detection of at least one triggering event, said triggering event comprising a change in any one of said respective input signals at a time, τ, greater-than a preselected amount indicating a condition deviation;
• wherein the step of operating further comprises:
• (a) upon said detection at said time, τ, a respective one of said learned output values, ONN, of said first set of data corresponding with said condition deviation is added to the proportional-plus-integral feedback controller current output, Oτ, and at said time, τ, a prior control output value, Oτ−1, of the proportional-plus-integral feedback controller associated with a prior time, τ−1, does not contribute to said current output Oτ;
• and
• (b) the proportional-plus-integral feedback controller functioning with a proportional gain constant, Kp, and an integral gain constant, Ki, that remain unchanged during the operating.