路由算法试图最小化成本 - 通过距离或行程时间来衡量。我天真地尝试将您的问题转化为这样的优化。我不确定以下想法是否有效：

假设每个旅行时间有多个POI（取最大值或高于它的值）。现在，当您计算细分市场的成本时，请使用该值与该细分市场中遇到的POI数量之间的差异作为错过的POI的度量。然后使用常用算法来最小化错过的POI数量。

我担心可能不会总是满足可用时间的约束。

Routing algorithms try to minimize costs - measured by the distance or by the travel time. I'd naively try to translate your problem into such an optimization. I am not sure whether following idea works:
Assume a number of POIs per travel time (take the maximum or a value above it). When you now calculate the costs of your segments, use the difference between that value and the number of POIs encountered along that segment as a measure for "missed POIs". And then use the common algorithms for minimizing the number of missed POIs.
I fear that the constraint of the available time might not always be met.