At pressent

the optimal load distribution system in thermal power plants is ussually based on the unit coal consumption characteristics

these charateristic curves

however

are valid only under stable conditions

the problem is that their corresponding results of "optimal" load distribution may be not true when the load of unit changes frequently.At the same time

the plant-level load distribution should also satisfy the adjustment time required under AGC mode.Directing against economy

rapidity

and stability needed in operation of units

a mathematic model of optimal load distribution has been established by using multi-objective fuzzy programming

and the immune algorithm being used to the calculation.The result obtained not only can ensure lower coal consumption

but also make the adjustment of unit load to be faster

at the same time

less units are needed to participate into adjustment

making the load distribution among units to be more resonable.