Traditionally, the easiest ways to separate water from low-level contaminants are (a) reverse osmosis, and (b) multi-effect distillation. 

Sir, Can this be done in a bottle like structure using little space ?

Reverse osmosis requires a semi-permeable membrane and a fairly high pressure pump. Otherwise, it is quite simple. Multi-effect distillation/evaporation requires a purpose-built machine which is a series of heat exchangers and evaporation chambers driven by a pump. Neither of these processes works well as a consumer product for the application you appear to have in mind.That is, in both cases you need an industrial or at least equipment-friendly setting. The armed services of many countries use the RO process (actually both processes, but most recently favor the RO process) for water purification quite routinely. 

I suspect the route you chose is not a good one.  Consider that some surfaces can cause the formation of liquid phase water from moist air.  That process requires no ongoing energy input.

Vacuum distillation is the commercial method for pure components. High temperature is avoided to prevent carbonization of glycerol. In waste water biological activity of the aerobic type quickly destroys the glycerol making carbon dioxide and water. 

Thermodynamics is the first thing you must consider to see if it is commercially viable.  You are adding a complex separation to a simple capture and the energy involved is critical.  Is it going to be more than the water is worth?  More than a more direct use of the energy, say with a cooled surface behind a membrane?

Article Thermodynamics properties of glycerol-water solution

I agree with Paul: probably some solid alternatives like CaO or some sorbents should be considered that can be just heated to a reasonable temperature to dehydrate or (unload) them. Once water ended up in glycerol it is not such a simple task to separate it as it might appear: boiling won't work, osmosis or vacuum evaporation are all rather industrial methods.

If one thinks along the lines of liquids, probably some concentrated caustic solutions could be used and then again the excessive water can be boiled out, but those mineral liquids, be it caustic or sulfuric acid, are pretty aggressive and may not be very user-safe in non-industrial applications.

Why dont you convert glycerol into glycerides that is insoluble in water 

Sir, My idea is to put all these processes in a bottle like structure . We are using glycerol as it absorbs water to a great extent .

 You may benefit from salting out effect or adding surface inactive material as potassium carbonate or sodium chloride to have completely immiscible mixture of water and glycerol then separate them by a separatory funnel then by applying reverse osmosis, you can get water.

Liquid-liquid extraction, perhaps by using tert-amyl alcohol, could possibly be considered to extract some of the glycerol from the aqueous phase.The ternary phase diagram (with tie lines) for the glycerol―water―tert-amyl alcohol can be found at: Alfred W. Francis, "Liquid-Liquid Equilibriums", Interscience Publishers (John Wiley & Sons), New York and London, 1963, p. 36.