算法描述
目前在物流，企业用工等领域，都有着大量的通过算法对接到的订单进行智能分配的需求。本文模拟的是用户下订单，然后商家接到订单，由配送人员进行派送的场景。在实际的应用中类似于百度外卖等有着非常多的实际应用。这种问题因为算法的复杂度太高，很难在短的时间周期内求解成功，所以有了像遗传算法，退火算法等启发式算法，以便在短的时间内能够求出近似的最优解。
本文模拟8个骑士，40个订单和40个商家进行计算。
算法描述
目前在物流，企业用工等领域，都有着大量的通过算法对接到的订单进行智能分配的需求。首先设计bean，商家接到订单的bean，这里需要商家和订单的坐标信息。
public class BusinessmenAndOrder {
private String id;//job标号
private String businessX;//商家坐标x
private String businessy;//商家坐标y
private String orderx;//订单坐标x
private String ordery;//订单坐标y
public BusinessmenAndOrder(){
}
public BusinessmenAndOrder(String id,String businessX,String businessy,String orderx,String ordery){
this.id=id;
this.businessX=businessX;
this.businessy=businessy;
this.orderx=orderx;
this.ordery=ordery;
}
public String getId() {
return id;
}
public void setId(String id) {
this.id = id;
}
public String getBusinessX() {
return businessX;
}
public void setBusinessX(String businessX) {
this.businessX = businessX;
}
public String getBusinessy() {
return businessy;
}
public void setBusinessy(String businessy) {
this.businessy = businessy;
}
public String getOrderx() {
return orderx;
}
public void setOrderx(String orderx) {
this.orderx = orderx;
}
public String getOrdery() {
return ordery;
}
public void setOrdery(String ordery) {
this.ordery = ordery;
}
}
模拟40个商家接到40个订单，其中商家和订单都可以是同一坐标位置，后续会考虑将坐标相近的位置进行聚类计算。
// 得到取件坐标和送达坐标（相当于商家和订单）
List<BusinessmenAndOrder> baoList = new ArrayList<BusinessmenAndOrder>();
BusinessmenAndOrder bao1 = new BusinessmenAndOrder("商家订单1", "137.45", "554.19", "138.83", "557.56");
BusinessmenAndOrder bao2 = new BusinessmenAndOrder("商家订单2", "840.6", "532.46", "139.37", "558.54");
BusinessmenAndOrder bao3 = new BusinessmenAndOrder("商家订单3", "157.38", "573.96", "359.96", "794.99");
BusinessmenAndOrder bao4 = new BusinessmenAndOrder("商家订单4", "120.45", "604.18", "193.49", "571.18");
BusinessmenAndOrder bao5 = new BusinessmenAndOrder("商家订单5", "137.50", "564.16", "357.33", "784.03");
BusinessmenAndOrder bao6 = new BusinessmenAndOrder("商家订单6", "131.90", "570.68", "118.60", "597.49");
BusinessmenAndOrder bao7 = new BusinessmenAndOrder("商家订单7", "150.21", "557.68", "140.98", "573.79");
BusinessmenAndOrder bao8 = new BusinessmenAndOrder("商家订单8", "115.13", "585.43", "115.01", "596.83");
BusinessmenAndOrder bao9 = new BusinessmenAndOrder("商家订单9", "120.13", "641.55", "141.20", "593.01");
BusinessmenAndOrder bao10 = new BusinessmenAndOrder("商家订单10", "150.06", "575.78", "355.11", "784.26");
BusinessmenAndOrder bao11 = new BusinessmenAndOrder("商家订单11", "114.50", "588.19", "119.16", "583.38");
BusinessmenAndOrder bao12 = new BusinessmenAndOrder("商家订单12", "137.28", "563.04", "355.24", "790.13");
BusinessmenAndOrder bao13 = new BusinessmenAndOrder("商家订单13", "151.04", "609.23", "179.37", "570.66");
BusinessmenAndOrder bao14 = new BusinessmenAndOrder("商家订单14", "113.73", "583.88", "139.83", "538.04");
BusinessmenAndOrder bao15 = new BusinessmenAndOrder("商家订单15", "185.75", "551.86", "187.18", "543.81");
BusinessmenAndOrder bao16 = new BusinessmenAndOrder("商家订单16", "116.39", "500.71", "173.63", "583.75");
BusinessmenAndOrder bao17 = new BusinessmenAndOrder("商家订单17", "112.73", "534.10", "108.60", "604.41");
BusinessmenAndOrder bao18 = new BusinessmenAndOrder("商家订单18", "137.66", "590.41", "139.63", "577.69");
BusinessmenAndOrder bao19 = new BusinessmenAndOrder("商家订单19", "161.38", "558.01", "161.83", "560.24");
BusinessmenAndOrder bao20 = new BusinessmenAndOrder("商家订单20", "137.28", "561.46", "143.01", "616.94");
BusinessmenAndOrder bao21 = new BusinessmenAndOrder("商家订单21", "128.48", "570.59", "135.86", "597.34");
BusinessmenAndOrder bao22 = new BusinessmenAndOrder("商家订单22", "136.45", "564.88", "139.55", "558.59");
BusinessmenAndOrder bao23 = new BusinessmenAndOrder("商家订单23", "128.96", "568.38", "139.37", "559.08");
BusinessmenAndOrder bao24 = new BusinessmenAndOrder("商家订单24", "133.41", "581.50", "357.09", "789.24");
BusinessmenAndOrder bao25 = new BusinessmenAndOrder("商家订单25", "134.83", "573.03", "136.14", "578.20");
BusinessmenAndOrder bao26 = new BusinessmenAndOrder("商家订单26", "146.76", "516.91", "165.01", "483.61");
BusinessmenAndOrder bao27 = new BusinessmenAndOrder("商家订单27", "127.41", "634.78", "152.30", "598.38");
BusinessmenAndOrder bao28 = new BusinessmenAndOrder("商家订单28", "137.48", "549.33", "141.20", "561.10");
BusinessmenAndOrder bao29 = new BusinessmenAndOrder("商家订单29", "178.75", "595.65", "204.53", "619.03");
BusinessmenAndOrder bao30 = new BusinessmenAndOrder("商家订单30", "97.72", "564.24", "640.4", "556.01");
BusinessmenAndOrder bao31 = new BusinessmenAndOrder("商家订单31", "628.48", "260.59", "125.86", "497.34");
BusinessmenAndOrder bao32 = new BusinessmenAndOrder("商家订单32", "336.45", "364.88", "439.55", "458.59");
BusinessmenAndOrder bao33 = new BusinessmenAndOrder("商家订单33", "228.96", "268.38", "239.37", "559.08");
BusinessmenAndOrder bao34 = new BusinessmenAndOrder("商家订单34", "153.41", "291.50", "257.09", "489.24");
BusinessmenAndOrder bao35 = new BusinessmenAndOrder("商家订单35", "136.83", "363.03", "276.14", "168.20");
BusinessmenAndOrder bao36 = new BusinessmenAndOrder("商家订单36", "446.76", "216.91", "465.01", "283.61");
BusinessmenAndOrder bao37 = new BusinessmenAndOrder("商家订单37", "527.41", "674.78", "158.30", "528.38");
BusinessmenAndOrder bao38 = new BusinessmenAndOrder("商家订单38", "637.48", "549.33", "191.20", "591.10");
BusinessmenAndOrder bao39 = new BusinessmenAndOrder("商家订单39", "637.48", "549.33", "291.20", "491.10");
BusinessmenAndOrder bao40 = new BusinessmenAndOrder("商家订单40", "637.48", "549.33", "691.20", "291.10");
然后将商家和订单加入到集合中。
baoList.add(bao1); baoList.add(bao2); baoList.add(bao3); baoList.add(bao4); baoList.add(bao5); baoList.add(bao6); baoList.add(bao7); baoList.add(bao8); baoList.add(bao9); baoList.add(bao10); baoList.add(bao11); baoList.add(bao12); baoList.add(bao13); baoList.add(bao14); baoList.add(bao15); baoList.add(bao16); baoList.add(bao17); baoList.add(bao18); baoList.add(bao19); baoList.add(bao20); baoList.add(bao21); baoList.add(bao22); baoList.add(bao23); baoList.add(bao24); baoList.add(bao25); baoList.add(bao26); baoList.add(bao27); baoList.add(bao28); baoList.add(bao29); baoList.add(bao30); baoList.add(bao31); baoList.add(bao32); baoList.add(bao33); baoList.add(bao34); baoList.add(bao35); baoList.add(bao36); baoList.add(bao37); baoList.add(bao38); baoList.add(bao39); baoList.add(bao40);
然后将商家和订单加入到集合中。
接着模拟骑士，假设骑士位置散落在不同的坐标点。
public class Driver {
private String driverId;//骑士编号
private String driverx;//骑士坐标x
private String drivery;//骑士坐标y
public Driver(){
}
public Driver(String driverId,String driverx,String drivery){
this.driverId=driverId;
this.driverx=driverx;
this.drivery=drivery;
}
public String getDriverId() {
return driverId;
}
public void setDriverId(String driverId) {
this.driverId = driverId;
}
public String getDriverx() {
return driverx;
}
public void setDriverx(String driverx) {
this.driverx = driverx;
}
public String getDrivery() {
return drivery;
}
public void setDrivery(String drivery) {
this.drivery = drivery;
}
}
模拟8个骑士
// 模拟骑士
List<Driver> driveList = new ArrayList<Driver>();
Driver driver1 = new Driver("driveA", "137.50", "575.78");
Driver driver2 = new Driver("driveB", "151.04", "534.10");
Driver driver3 = new Driver("driveC", "137.28", "558.01");
Driver driver4 = new Driver("driveD", "127.41", "634.78");
Driver driver5 = new Driver("driveE", "46.76", "285.75");
Driver driver6 = new Driver("driveF", "176.76", "485.75");
Driver driver7 = new Driver("driveG", "386.76", "405.75");
Driver driver8 = new Driver("driveH", "246.76", "685.75");
driveList.add(driver1);
driveList.add(driver2);
driveList.add(driver3);
driveList.add(driver4);
driveList.add(driver5);
driveList.add(driver6);
driveList.add(driver7);
driveList.add(driver8);
将模拟数据代入算法
List<DistributionSep> distList = new ArrayList<DistributionSep>();
distList = autoSeperateOrder(baoList, driveList);
for (DistributionSep distribution : distList) {
System.out.println("配送人员 "+distribution.getDriverName());
List<OrderSort> orderSet = distribution.getDist();
for (OrderSort order : orderSet) {
System.out.println(order.getOrder() + " " + order.getPors() + " " + order.getBussinessOrder());
}
}
结果查看
查看运行结果，可知，迭代了1024次，耗时10秒。
2017-06-30 14:49:38,888 INFO [main] problem.VehicleRoutingProblem - setup problem: [fleetSize=FINITE][#jobs=40][#vehicles=8][#vehicleTypes=1][transportCost=[name=crowFlyCosts]][activityCosts=com.graphhopper.jsprit.core.problem.cost.WaitingTimeCosts@47d90b9e] 2017-06-30 14:49:38,970 INFO [main] algorithm.VehicleRoutingAlgorithm - algorithm starts: [maxIterations=2000] 2017-06-30 14:49:38,972 INFO [main] algorithm.InsertionInitialSolutionFactory - create initial solution 2017-06-30 14:49:39,823 INFO [main] algorithm.VehicleRoutingAlgorithm - iterations start 2017-06-30 14:49:39,825 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 1 2017-06-30 14:49:39,843 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 2 2017-06-30 14:49:39,910 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 4 2017-06-30 14:49:40,032 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 8 2017-06-30 14:49:40,142 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 16 2017-06-30 14:49:41,014 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 32 2017-06-30 14:49:41,194 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 64 2017-06-30 14:49:41,728 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 128 2017-06-30 14:49:42,439 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 256 2017-06-30 14:49:43,599 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 512 2017-06-30 14:49:45,723 INFO [main] algorithm.VehicleRoutingAlgorithm$Counter - iterations 1024 2017-06-30 14:49:49,703 INFO [main] algorithm.VehicleRoutingAlgorithm - iterations end at 2000 iterations 2017-06-30 14:49:49,703 INFO [main] algorithm.VehicleRoutingAlgorithm - took 10.732 seconds
分配的结果及cost计算结果为：
+--------------------------------------------------------------------------------------------------------------------------------+ | detailed solution | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | route | vehicle | activity | job | arrTime | endTime | costs | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 1 | driveA | start | - | undef | 0 | 0 | | 1 | driveA | pickupShipment | 商家订单25 | 4 | 4 | 4 | | 1 | driveA | deliverShipment | 商家订单25 | 9 | 9 | 9 | | 1 | driveA | pickupShipment | 商家订单21 | 20 | 20 | 20 | | 1 | driveA | pickupShipment | 商家订单23 | 22 | 22 | 22 | | 1 | driveA | pickupShipment | 商家订单22 | 30 | 30 | 30 | | 1 | driveA | pickupShipment | 商家订单20 | 34 | 34 | 34 | | 1 | driveA | deliverShipment | 商家订单22 | 38 | 38 | 38 | | 1 | driveA | deliverShipment | 商家订单23 | 38 | 38 | 38 | | 1 | driveA | pickupShipment | 商家订单18 | 70 | 70 | 70 | | 1 | driveA | deliverShipment | 商家订单18 | 82 | 82 | 82 | | 1 | driveA | deliverShipment | 商家订单21 | 102 | 102 | 102 | | 1 | driveA | deliverShipment | 商家订单20 | 123 | 123 | 123 | | 1 | driveA | end | - | 123 | undef | 123 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 2 | driveB | start | - | undef | 0 | 0 | | 2 | driveB | pickupShipment | 商家订单7 | 24 | 24 | 24 | | 2 | driveB | deliverShipment | 商家订单7 | 42 | 42 | 42 | | 2 | driveB | pickupShipment | 商家订单19 | 68 | 68 | 68 | | 2 | driveB | deliverShipment | 商家订单19 | 70 | 70 | 70 | | 2 | driveB | pickupShipment | 商家订单15 | 96 | 96 | 96 | | 2 | driveB | deliverShipment | 商家订单15 | 104 | 104 | 104 | | 2 | driveB | end | - | 104 | undef | 104 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 3 | driveC | start | - | undef | 0 | 0 | | 3 | driveC | pickupShipment | 商家订单1 | 4 | 4 | 4 | | 3 | driveC | pickupShipment | 商家订单28 | 9 | 9 | 9 | | 3 | driveC | deliverShipment | 商家订单1 | 17 | 17 | 17 | | 3 | driveC | deliverShipment | 商家订单28 | 21 | 21 | 21 | | 3 | driveC | pickupShipment | 商家订单12 | 26 | 26 | 26 | | 3 | driveC | pickupShipment | 商家订单5 | 27 | 27 | 27 | | 3 | driveC | pickupShipment | 商家订单6 | 35 | 35 | 35 | | 3 | driveC | pickupShipment | 商家订单8 | 58 | 58 | 58 | | 3 | driveC | pickupShipment | 商家订单11 | 61 | 61 | 61 | | 3 | driveC | deliverShipment | 商家订单8 | 69 | 69 | 69 | | 3 | driveC | deliverShipment | 商家订单6 | 73 | 73 | 73 | | 3 | driveC | deliverShipment | 商家订单11 | 87 | 87 | 87 | | 3 | driveC | pickupShipment | 商家订单24 | 101 | 101 | 101 | | 3 | driveC | pickupShipment | 商家订单10 | 119 | 119 | 119 | | 3 | driveC | pickupShipment | 商家订单3 | 127 | 127 | 127 | | 3 | driveC | pickupShipment | 商家订单29 | 157 | 157 | 157 | | 3 | driveC | deliverShipment | 商家订单29 | 192 | 192 | 192 | | 3 | driveC | deliverShipment | 商家订单10 | 415 | 415 | 415 | | 3 | driveC | deliverShipment | 商家订单5 | 418 | 418 | 418 | | 3 | driveC | deliverShipment | 商家订单24 | 423 | 423 | 423 | | 3 | driveC | deliverShipment | 商家订单12 | 425 | 425 | 425 | | 3 | driveC | deliverShipment | 商家订单3 | 432 | 432 | 432 | | 3 | driveC | end | - | 432 | undef | 432 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 4 | driveD | start | - | undef | 0 | 0 | | 4 | driveD | pickupShipment | 商家订单27 | 0 | 0 | 0 | | 4 | driveD | pickupShipment | 商家订单9 | 10 | 10 | 10 | | 4 | driveD | pickupShipment | 商家订单13 | 55 | 55 | 55 | | 4 | driveD | deliverShipment | 商家订单27 | 66 | 66 | 66 | | 4 | driveD | deliverShipment | 商家订单9 | 78 | 78 | 78 | | 4 | driveD | pickupShipment | 商家订单14 | 107 | 107 | 107 | | 4 | driveD | pickupShipment | 商家订单30 | 132 | 132 | 132 | | 4 | driveD | deliverShipment | 商家订单14 | 182 | 182 | 182 | | 4 | driveD | deliverShipment | 商家订单13 | 233 | 233 | 233 | | 4 | driveD | pickupShipment | 商家订单37 | 596 | 596 | 596 | | 4 | driveD | deliverShipment | 商家订单30 | 760 | 760 | 760 | | 4 | driveD | pickupShipment | 商家订单40 | 768 | 768 | 768 | | 4 | driveD | pickupShipment | 商家订单39 | 768 | 768 | 768 | | 4 | driveD | pickupShipment | 商家订单38 | 768 | 768 | 768 | | 4 | driveD | pickupShipment | 商家订单2 | 971 | 971 | 971 | | 4 | driveD | deliverShipment | 商家订单40 | 1255 | 1255 | 1255 | | 4 | driveD | pickupShipment | 商家订单31 | 1325 | 1325 | 1325 | | 4 | driveD | deliverShipment | 商家订单39 | 1734 | 1734 | 1734 | | 4 | driveD | deliverShipment | 商家订单38 | 1875 | 1875 | 1875 | | 4 | driveD | deliverShipment | 商家订单2 | 1936 | 1936 | 1936 | | 4 | driveD | deliverShipment | 商家订单37 | 1972 | 1972 | 1972 | | 4 | driveD | deliverShipment | 商家订单31 | 2017 | 2017 | 2017 | | 4 | driveD | end | - | 2017 | undef | 2017 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 5 | driveE | start | - | undef | 0 | 0 | | 5 | driveE | pickupShipment | 商家订单35 | 119 | 119 | 119 | | 5 | driveE | pickupShipment | 商家订单34 | 192 | 192 | 192 | | 5 | driveE | pickupShipment | 商家订单33 | 271 | 271 | 271 | | 5 | driveE | deliverShipment | 商家订单35 | 382 | 382 | 382 | | 5 | driveE | pickupShipment | 商家订单36 | 559 | 559 | 559 | | 5 | driveE | deliverShipment | 商家订单36 | 628 | 628 | 628 | | 5 | driveE | deliverShipment | 商家订单34 | 921 | 921 | 921 | | 5 | driveE | deliverShipment | 商家订单33 | 993 | 993 | 993 | | 5 | driveE | end | - | 993 | undef | 993 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 6 | driveF | start | - | undef | 0 | 0 | | 6 | driveF | pickupShipment | 商家订单26 | 43 | 43 | 43 | | 6 | driveF | deliverShipment | 商家订单26 | 81 | 81 | 81 | | 6 | driveF | pickupShipment | 商家订单16 | 133 | 133 | 133 | | 6 | driveF | pickupShipment | 商家订单17 | 166 | 166 | 166 | | 6 | driveF | deliverShipment | 商家订单17 | 237 | 237 | 237 | | 6 | driveF | pickupShipment | 商家订单4 | 249 | 249 | 249 | | 6 | driveF | deliverShipment | 商家订单16 | 306 | 306 | 306 | | 6 | driveF | deliverShipment | 商家订单4 | 329 | 329 | 329 | | 6 | driveF | end | - | 329 | undef | 329 | +---------+----------------------+-----------------------+-----------------+-----------------+-----------------+-----------------+ | 7 | driveG | start | - | undef | 0 | 0 | | 7 | driveG | pickupShipment | 商家订单32 | 65 | 65 | 65 | | 7 | driveG | deliverShipment | 商家订单32 | 204 | 204 | 204 | | 7 | driveG | end | - | 204 | undef | 204 | +--------------------------------------------------------------------------------------------------------------------------------+
这样就可以得到
配送人员 driveA 1 取 商家订单25 2 送 商家订单25 3 取 商家订单21 4 取 商家订单23 5 取 商家订单22 6 取 商家订单20 7 送 商家订单22 8 送 商家订单23 9 取 商家订单18 10 送 商家订单18 11 送 商家订单21 12 送 商家订单20 配送人员 driveB 1 取 商家订单7 2 送 商家订单7 3 取 商家订单19 4 送 商家订单19 5 取 商家订单15 6 送 商家订单15 配送人员 driveC 1 取 商家订单1 2 取 商家订单28 3 送 商家订单1 4 送 商家订单28 5 取 商家订单12 6 取 商家订单5 7 取 商家订单6 8 取 商家订单8 9 取 商家订单11 10 送 商家订单8 11 送 商家订单6 12 送 商家订单11 13 取 商家订单24 14 取 商家订单10 15 取 商家订单3 16 取 商家订单29 17 送 商家订单29 18 送 商家订单10 19 送 商家订单5 20 送 商家订单24 21 送 商家订单12 22 送 商家订单3 配送人员 driveD 1 取 商家订单27 2 取 商家订单9 3 取 商家订单13 4 送 商家订单27 5 送 商家订单9 6 取 商家订单14 7 取 商家订单30 8 送 商家订单14 9 送 商家订单13 10 取 商家订单37 11 送 商家订单30 12 取 商家订单40 13 取 商家订单39 14 取 商家订单38 15 取 商家订单2 16 送 商家订单40 17 取 商家订单31 18 送 商家订单39 19 送 商家订单38 20 送 商家订单2 21 送 商家订单37 22 送 商家订单31 配送人员 driveE 1 取 商家订单35 2 取 商家订单34 3 取 商家订单33 4 送 商家订单35 5 取 商家订单36 6 送 商家订单36 7 送 商家订单34 8 送 商家订单33 配送人员 driveF 1 取 商家订单26 2 送 商家订单26 3 取 商家订单16 4 取 商家订单17 5 送 商家订单17 6 取 商家订单4 7 送 商家订单16 8 送 商家订单4 配送人员 driveG 1 取 商家订单32 2 送 商家订单32
配送人员的配送顺序。
总结
目前算法中使用的是欧式距离来进行两个点之间距离的计算，运算的时间随着单子数量的增加会加长，并且在某些情况下还会有很多的约束条件，这些都是下一步算法需要改进的重点。