Optaplanner VRP增量分数过约束规划

Question

我想为VRP创建一个过度约束规划的增量分数。我创建了一个传统的虚拟车辆，其中包括所有计划外的客户。

示例：

Optaplanner将Customer1从Vehicle1移动到Vehicle2:

beforeVariableChanged:  previousStandstill(Customer1),  customer.GetVehicle() = Vehicle1
beforeVariableChanged:  nextCustomer(Customer0),        customer.GetVehicle() = Vehicle1
afterVariableChanged:   nextCustomer(Customer0),        customer.GetVehicle() = Vehicle1
afterVariableChanged:   previousStandstill(Customer1),  customer.GetVehicle() = Vehicle1
beforeVariableChanged:  vehicle(Customer1),             customer.GetVehicle() = Vehicle1
afterVariableChanged:   vehicle(Customer1),             customer.GetVehicle() = Vehicle2

当我得到AfterVariableChanged：previousStandstill(Customer1)时，在Customer.getVehicle（）中，我有旧车辆的价值，我不知道是否需要添加软成本（对于虚拟车辆的成本被忽略）。有没有办法在AfterVariableChanged-PreviousStandstill而不是AfterVariableChanged-Vehicle中获得实际车辆。

Answer 1

检查是否使用AnchorShadowVariable注释了vechicle字段：

@AnchorShadowVariable(sourceVariableName = "previousStandstill")
public VehicleNode getVehicleNode() {
    return vehicleNode;
}

该注释告诉Optaplanner更新vechicle字段。