为了确定何时做出动作，我们将观察动作规划中第一步动作是否足够大，或者前三步动作之和是否够大（加速动作），或者我们积累但未执行的多回合小动作累积（MOVACC）超过MOVRES时。
需要对MV的ramp操作进行特殊考虑。
该功能不推荐MVs作为主要ramp操作。Ramp将不会得到较好的控制。动态动作规划将经常更改MVs以将ramp移至周围（即使MV的SS目标并未改变）。该功能将会频繁地抑制或加速动作以控制ramp，因此ramp控制将不会很好。要注意这个问题。
稳态求解器
稳态求解器必须提供一个动态动作规划可以满足的，由最小tuning动作带来新约束的解决方案。这是非常重要的，因此确定稳态解决方案时必须确保目标中小的动作被拒绝。
一个迭代方法被用来验证并解决方案。若有需要，可通过调整有效操作限来限定动作幅度，防止没有任何动作。
有一个例外是VIND超出限制。用户改变约束时会导致这一情况。例如当动作小于MOVRES时会发生这种情况。
通过运用最低动作约束改良稳态目标变量的稳态解决方案指标MANACT=6 (SS MINMOV)。
一些后果：
•对于单个MOVRES变量，当一个小的动作被拒绝时你可能会观察到MANACT的SS MINMOV。
•对于单个MOVRES变量，当一个中等大小的动作被提升为全尺寸动作时你可能会观察到MANACT的SS MINMOV。
•当MV开始远离它的目标时（比如你突然显著扩大限制），由于SS求解器将拒绝进一步动作，最终MV将不会触碰到限制并且会接近目标值。
操作员可能需要向限制中添加一点缓冲以使其获得真正想要的动作。
•对于多个MOVRES变量，首先我们拒绝相对最小的动作，并期望LP依旧能找到一个较好的解决方案。为弥补被拒绝的动作，其可能要移动更多其它MINMOV变量。
附原文：
In order to determine when to make a move, we look to see if the first move in the move plan is big enough OR the first three moves combined are big enough(accelerate the moves) OR we have accumulated, but not implemented, small moves for some number of cycles and our accumulation (MOVACC) has exceeded the MOVRES value.
Special consideration for ramp handle manipulated variables.
This feature is not recommended for MVs that are primary ramp handles. The ramp will not be controlled very well. The dynamic move plan will often make changes to MVs to move the ramps around (even if there is no change in SS target for the MV). This feature will frequently repress or accelerate the moves used to control the ramp, thus the control of the ramp is not very good. Be aware ofthis problem.
Steady state solver
The steady state solver must provide a solution that the dynamic move plan can meet, given the new constraint imposed by the minimum movement tuning. It is important, therefore, to make sure that small moves in the target are rejected when validating the steady state solution.
An iterative method is used to validate the solution and then resolve. If necessary, the effective operating limits are adjusted to force a move of a certain size or to prevent any movement at all.
An exception is when the VIND is outside limits. The user causes this situation by moving the limits. This is a situation where a move less than MOVRES can be implemented.
Variables that have a modified steady state target to meet the minimum movement constraint will have a steady state solution indicator of MANACT=6 (SS MINMOV).
Some ramifications:
·With a single MOVRES variable, you will likely see SS MINMOV for the MANACT whenever a small move is rejected.
·With a single MOVRES variable, you will sometime see SS MINMOV for the MANACT when a medium sized move is promoted to a full size move.
·When a MV is far away from its target (say you suddenly and significantly widen the limits), then you will eventually get close enough where the SS solver will reject any further moves,and you won’t be able to get to the limit.
Operators may need to add a little cushion to the limits to get the moves they really want (best practices).
~With multiple MOVRES variables, we reject the smallest relative moves first, in hopes that the LP will still find a pretty good solution. It might even move other MINMOV variables more to compensate for the rejected moves.
2015.9.24
