Control chain principles

Control is a common-sense concept. Examples can be found in the natural world, including the living world (e.g. temperature regulation in the human body). Its first technological achievements can be traced back to antiquity (Ktesibios clepsydra in – 300: a level regulator (like a water flush) maintains a constant flow at the outlet of a reservoir; this flow feeds a tank whose height is then used to measure time). But it wasn't until the 19th century that the properties of control loops were formally studied. Since then, progress has been constant.

As an effective response to tough industrial competition, automation has undergone major changes in recent years. The classic sectors that saw the birth of modern automation, such as petrochemicals, iron and steel, nuclear power and energy production, as well as the more recent sectors of pharmaceuticals and agri-food, are still fields of application for advanced control. But other sectors have recently opened up, in industry (space, transport), research (giant telescopes, space probes, sophisticated medical equipment, autonomous vehicles) or everyday life (laser disk reading heads, "intelligent" household appliances).

Automation can serve a variety of purposes. For example:

• increase the performance of the production system: improve the quality of the finished product, reduce its variability, increase the quantity of product manufactured or reduce manufacturing time;

• Reduce manufacturing costs: minimize the quantity of raw materials used, the energy required for the manufacturing process, introduce by-product recovery (recycling of which leads to internal loops and strong coupling between variables);

• improve the safety of the industrial facility, the people who work there and the environment; enable product traceability (a necessity in the agri-food industry, for example);

• remote control: this is particularly important for nuclear and space applications;

• offer personalized products and services, improve comfort (e.g. in the automotive industry, where vehicle models are modular and more and more variables are controlled – ABS braking, brake force distribution, rain sensor, automatic low beam headlights, etc.);

• reduce the time needed for innovation.

The variety of problems naturally leads to the variety of solutions. In addition, we note that the automation of a plant generally leads to the reorganization of production, the rationalization of operations, and synchronization between various workshops, between production, supply and demand. Indeed, the global optimum is not...