Apparatus for controlling glass melting and/or refining furnaces

A system for controlling the melting of a glass batch in a glass melting furnace and comprising:

• a plurality of sensors for detecting different types of operating conditions in a furnace including;
• temperature information obtained from temperature sensors selectively set into or on the glass melting furnace;
• information relating to flow rates and pressures of fluids used by the furnace including:
• fuel, oxidizer, fume as well as their compositions;
• cooling parameters, electricity;
• and the measurements of the consumption of each fluid, this information being delivered by sensors selectively provided in or on the furnace;
• information regarding position of the various furnace actuators including:
• control valves, devices for varying electrical power, and batch flow rate;
• information from end-of-travel sensors for the furnace;
• set point values which are manually input by the operators into the furnace system;
• information relating to quality of glass produced;
• information from a neural network;
• and
• information coming from analysis of an image of the inside of the furnace;
• means for creating, and analyzing images taken inside a furnace in accordance with a predetermined mathematical model;
• a predictive network which, depending on the state of the furnace and information regarding changes in production over time, defines various set point values assigned to a plurality of furnace actuators of different types that affect a plurality of furnace operations;
• means for storing operator set points corresponding to manual operation of furnace actuators under preselected conditions;
• fuzzy logic controller means running a fuzzy logic control algorithm and connected at a plurality of respective inputs to the sensors, image means, predictive network, and operator set points storing means, the controller means generating a plurality of output signals for respective actuators that will control melting in the furnace including:
• combustion equipment including burners with their fuel and oxidizer feed devices;
• electrical equipment for heating;
• glass refining devices including bubblers and boost melters;
• and
• wherein the furnace system operates in accordance with objectives defined by the operator set points;
• and
• the predictive network further including means for learning operating laws of the furnace during a learning phase, wherein the learning means defines the laws of different types of furnace operations, selectively from actual furnace operation, or by simulation of furnace operation using a mathematical model.