JP2000112925A - Method/system for generating delivery route for miniaturizing number of routes - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently obtain a delivery route with the small number of routes while various restriction conditions are satisfied for executing delivery through the use of plural routes. SOLUTION: The best solution (b) and a present solution (s) are set to initial values and are stored. Then, the present solution (s) is inputted to a reference solution, and a judgement solution to an evaluation function F as the solution near the present solution (s) so as to calculate an evaluation value. Thus, local search is executed and a local solution (t) is obtained. Then, the best solution (b) is inputted to the reference solution and the judgement solution to the evaluation function F as the local solution (t) so as to calculate the evaluation value. When the local solution (t) is better than the best solution (b), the local solution (t) is substituted for the best solution (b). When a termination condition which is previously decided is satisfied, a processing is terminated. When the termination condition is not satisfied, the best solution (b) is substituted for the present solution (s), the value of a correction flag is changed and local search and the substitution of the best solution are repeated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and a system for generating a delivery route for minimizing the number of routes, and more particularly to efficiently providing a delivery route with a small number of routes while satisfying a constraint condition for delivery by a plurality of routes. It is an invention relating to a method and a system for obtaining it.

[0002]

2. Description of the Related Art In the delivery of resources in the distribution industry and the like, how to efficiently deliver resources is one of the most fundamental and important issues. Although it is relatively easy to realize an approximate solution for minimizing the path length, it is almost always desirable to minimize the number of paths (minimize the number of vehicles used) in practical problems. For example, the problem of a shipping company delivering multiple deliveries to multiple destinations using fewer vehicles so as to satisfy various constraints (time, number of vehicles, vehicle carrying capacity, etc.) is a problem. Hit. 1 route
Then, it seems that there is no problem because one car is enough, but it takes time to complete the delivery, or it exceeds the carrying capacity of one car, etc. It is rare that one delivery vehicle is enough. The present invention is an invention that provides a method and a system for efficiently finding a delivery route with a small number of routes while satisfying various constraints when delivering using a plurality of routes.

FIG. 1 shows an example of a delivery route. A delivery destination (node) is represented by a circle, and a point serving as a delivery base is represented by a double circle. Here, the route means a route on which a certain delivery vehicle starts from a base and returns to the base again after the material is delivered. Since the route is determined for each vehicle, the number of routes is equal to the number of vehicles. The size of the delivery amount delivered to the delivery destination is represented by the size of a circle. Although the number of bases is one in FIG. 1 for convenience of explanation, there is no limitation on the number of bases in the embodiment of the present invention.

FIG. 2 shows an example of a method of expressing a route. Each distribution
Numbers (1,2,3,…, 70) are assigned to destinations, and routed through each route
Expressed as a transit point array that lists the destination numbers of the destination. Each sutra
Each road is assigned a serial number i.
Length l_i, Constraint violation count p _i, Whether the route is empty
Gr_i, A flag m indicating whether to modify the objective function_iIs included
Suppose you are. The number of violations (penalty)
Means that certain routes are restricted (vehicle capacity constraints, travel time constraints,
(For example, when you stop by)
Value. The route length is usually the time required for delivery (running
Time and stop time). For example, Route 1
In the case of (i = 1), the delivery route is 30-> 64-> 27->
It goes around the delivery destination in order of 67-> 68-> 16. That time
Road length (time required for delivery) l₁Is 100.

Now, an initial solution (initial route) in which the number of paths is not minimized or does not satisfy the constraints is given, and the solution is updated so as to minimize the path length or satisfy the constraints. Local Search (LS)
(Assuming that it is not given, since it is difficult to implement a local search that updates the solution so as to minimize the number of paths). To minimize the number of paths, iterative local search (ILS: Iter
ated Local Search). FIG. 9 shows pseudo code of the iterative local search method. In the iterative local search method, escape from the local solution is performed because the local search must be repeatedly used.
In FIG. 9, s, t, b is a solution (b is the best solution), and n is an integer variable. ESC is an escape method from local solutions. If the solution changes completely at random, it is difficult to find a good solution in the next local search, so it is necessary to find a somewhat good escape solution.

[0006] There are the following known methods for escaping from a local solution in the iterative local search method. -Node sampling method: Randomly sample n nodes (n is a variable in the ILS) and insert them into the best positions. -Route extraction method: Several routes are randomly extracted, and their nodes are inserted into the best positions.

[0007] Even if the purpose of optimization is to minimize the number of paths,
Since the neighborhood where the number of routes decreases cannot be searched locally,
The local search will work for the purpose of time minimization. Therefore, even if the number of routes is several steps away from the vicinity, it may be caught in the vicinity of the point of time decreasing, and a local solution that does not decrease the number of paths may be reached. Escape methods must be designed to compensate for this drawback. However, in the above-described escape method, when inserting the extracted nodes into the best positions, the remaining paths are not touched, so that there is a low possibility of finding a good solution with a reduced number of paths. In other words, even if nodes can be inserted such that the number of routes is reduced by exchanging and moving partial routes between the remaining routes, this cannot be done.

[0008]

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a method for efficiently obtaining a delivery route having a small number of routes while satisfying various constraints in delivery using a plurality of routes. And a system. Yet another object is to provide a method and system for calculating a solution for a delivery route that minimizes the number of routes. Yet another object is to provide a new escape method from the local solution of the iterative local search method. Still another object is to provide a method and a system for escaping from a local solution, in which, in a repeated local search, nodes are inserted so that the number of routes is reduced if routes are exchanged and moved between routes to reduce the number of routes. That is.

[0009]

In order to solve the above-mentioned problem, first, a reference solution and a judgment solution are inputted, and a total path length function L,
A constraint function P, a non-empty path number function R, and a correction flag for correcting a function value for each path to be calculated by at least one of the L, P, and R functions;
An evaluation function F for calculating an evaluation result is prepared, and a method of calculating a solution of a delivery route that minimizes the number of routes is obtained by repeatedly using this function. In this method, first, the best solution b and the current solution s are set as initial values and stored, and then the reference solution is input to the evaluation function F as the current solution s and the judgment solution is a solution near the current solution s. , An evaluation value is calculated, a local search is performed, and a local solution t is obtained. Then, the evaluation function F is defined by setting the reference solution as the best solution b and the decision solution as the local solution t.
And the evaluation value is calculated, and the local solution t becomes the best solution b
If the solution is better than the best solution b, the local solution t
Is assigned. Here, if the predetermined termination condition is satisfied, the process is terminated. If the termination condition is not satisfied, the best solution b is substituted for the current solution s, and the value of the correction flag is changed. The local search and the replacement of the best solution are repeated. A feature of the present invention resides in that the objective function (L, P, R) is modified (that is, the modification flag is modified) in the local search. For example, a flag for correcting the path length is provided for each path, and the objective function L can be calculated with reference to the flag. As a simpler example, the number of routes can be minimized by the following procedure. First, a local search is performed until the solution cannot be updated, and a solution (local solution) that cannot be further updated by the local search is found. Next, the path lengths of some of the local solutions are long estimated, and the local search is executed again. For a newly generated route during the execution of the local search, the route length is estimated long. It is configured to output the best solution found while repeating the above operations several times.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention is a new method relating to a method for escaping from a local solution in a solution called an iterative local search method. FIG. 8 shows a definition example of the evaluation function of the present invention used in the iterative local search method. Let s be the current solution, and let N (s) be the neighborhood of s generated by the local search method LS. LS is the evaluation function F (s, N (s)) = {Y, N}. In the case of, the neighborhood is adopted (s <-N (s)). Evaluation function F
Is defined by three objective functions: L: Total travel time including total stop time (or total route length) P: Numerical value of the violated constraint R: Number of routes (number of vehicles used)

Here, the entire program for obtaining the delivery route is a flowchart as shown in FIG. At this time,
The solutions s, t, b, etc. respectively correspond to the waypoint arrays. In step 1 of FIG. 3, an appropriate initial solution is determined and set to internal variables (intermediate location arrays) b and s. In step 2 of FIG. 3, a process of gradually improving the solution is performed while searching for the vicinity of a certain solution. FIG. 4 is a flowchart showing this processing in more detail. First, the neighborhood of the current solution s N (s)
For each solution u∈N (s), the objective function L (u),
Calculate P (u) and R (u), if F (s, u) is Y, substitute u for s, determine if s has been updated, and if so, repeat the same process again If s is not updated, s is substituted into t as a local solution.

The neighborhood function N (s) is a function that gives a set of solutions obtained by locally changing a certain solution s. For example, as shown in FIG. 12, a solution u obtained by exchanging the positions of two points is N (s)
Is included. When two places (delivery destinations) are exchanged,
FIGS. 13 and 14 show examples in which two consecutive transit points (delivery destinations) are replaced by transit point arrangements. Usually, of the solutions obtained by such replacement,
N (s) includes such a solution because it is easy to find only those that reduce the objective function L or P.
It should be noted that three or more points or three or more consecutive transit values may be exchanged.

The objective functions L, P, R will be described. The delivery destination is represented by j = 1, 2,... N, and the delivery base is represented by 0. E _i a set of moving sections (sides) included in a certain path i, a set of destinations that are included in a certain path i and V _i. For example, E _i
The {(0,30), (30,64), (64,27), ... (16, 0)} set of pairs of points, such as, V _i is {30, 64, 27, ... 16 }. The moving distance (time) between two points a and b is d _{(a, b)} , the working time at each delivery destination j is w _j , and the number of articles delivered to _j is q _j . M is the total number of routes including empty routes
And For example, M corresponds to the number of vehicles owned. Also,
For convenience, a function D that takes 1 if a certain condition c is true and 0 if it is false
Assume (c).

The objective function L is the total route length or the total delivery time,

(Equation 1) Is defined as Here, when L is the total path length, l _i is the path length of each path i,

(Equation 2) (At this time, d is the moving distance). When L is the total delivery time, l _i is the delivery time of each route i,

(Equation 3) (At this time, d is the moving time).

The objective function P is defined depending on the constraints to be considered. Here, as an example, a case is shown in which the total number of violations is set to P in consideration of the two constraint conditions of the maximum route length (delivery time) L _max and the load capacity Q _max . Such an objective function P is

(Equation 4) Is defined as Here, p _i is the number of violations in a certain route i,

(Equation 5) Given by

The objective function R is the total number of non-empty paths,

(Equation 6) Is defined as Here, r _i is a value that takes 1 if a certain route i is not empty, and takes 0 if it is empty.

(Equation 7) Given by Note that the objective functions L, P, and R are
A flag is provided for modifying the objective function for each path. This flag allows the value of each objective function to be modified.

For example, as a simple example of modifying the objective function L, L is l _i × (1 + m _i × C) for each path.
, P is defined as the sum of p _i for each path, and R is defined as the sum of r _i for each path. At this time, C is a constant.

Next, the evaluation function F will be described. The evaluation function F refers to the three objective functions L, P, and R to determine whether a certain solution can be compared with another solution.
If not, N is output. In the evaluation function F (s, t), adopting the decision solution t instead of the reference solution s can be achieved by (1) reducing the constraint P as much as possible, and (2) making the number of paths R as small as possible.
And (3) evaluate whether it is effective to reduce the total (L + P) of the path length and the constraint as much as possible (L, P, R ≧ 0). In the flowchart, the priorities are made higher in the order of the above numbers (that is, the lowering of the constraint condition is the highest priority), and (1) and (3) are simultaneously taken into account in order to avoid sacrificing L to lower P. An example is shown in FIG. The flowchart of FIG. 5 outputs Y under the following conditions. (A) If P (t) = 0 and R (s)> R (t), Y (b) P (s)> P (t) and L (s) + P (s)> L (t) If P (t), then Y (c) if R (s) = R (t) and L (s) + P (s)> L (t)> P (t), then Y above (a), If conditions (b) and (c) are not met, N
Is output. The outputs Y and N of the above evaluation function may be replaced with 1, 0, respectively.

FIG. 6 shows a flowchart describing step 3 of FIG. 3 in more detail. In step 3, it is determined whether the obtained local solution t can be compared with the best solution b obtained so far, and if it is good, it is stored as b. In step 4 of FIG. 3, it is determined whether or not to perform a further local search. This is performed by designating the number of executions (the number of repetitions) of the local search. Step 5
Then, the objective function which is the main subject of the present invention is corrected. Performing step 5, setting a flag for correcting the objective function for each path, and calculating the objective function with reference to this flag are features of the present invention.

FIG. 7 shows a flowchart of the object correction in the present invention. An appropriate constant is designated as the variable k described in FIG. 7 (for example, k = 2). First, the best solution b is substituted for the current solution s. For all of the route the correction flag m _i is set to 0. Then randomly k the chosen route non-empty, is set to 1 the correction flag m _i for each. Then, the correction flags _mi are set to 1 for all the empty routes.

In FIG. 10, using two actual data A and B,
The experimental results of evaluating the node extraction method, route extraction method, and purpose correction method are shown. The input data A is composed of 21 destinations and one base. Due to complicated time frame conditions, minimizing the number of routes is a difficult problem that cannot reduce R unless the value of L is considerably worse (within P = 0). The input data B includes 332 delivery destinations and 4 bases. For this problem, the difficulty of minimizing time and minimizing the number of routes remains the same. Using three escape methods, we performed an ILS with a maximum of 50 iterations and compared the number of iterations I until the best solution was obtained, the number of iterations I _R until the best R was obtained, and the best solution . Since both results are P = 0,
This is omitted in FIG. From FIG. 10, it can be seen that the objective correction method obtains a good solution with a smaller number of iterations compared to the other two methods. In particular, in data A, a solution with three paths, which was not obtained by the other two methods, was obtained.

As another embodiment, the present invention can be used for generating a route for a cash transport vehicle for exchanging ATM cassettes installed outside a bank store. The sum of the running time and ATM cassette exchange time of the First a cash transport vehicles i and l _i, the time required that the sum of the l _i for each transport vehicle taking into account the purpose Modify variable m _i L, violations of constraints Where P is the number of vehicles transported and R is the number of vehicles transported by cash, the number of vehicles transported by cash can be minimized by directly applying the solution according to the present invention. However, conditions such as the number of cassettes that can be loaded on the cash transport vehicle and the working hours of the driver are added to the constraint conditions.

In yet another embodiment, the present invention can be used to create a route for traveling around a parts factory and transporting parts to an assembly factory. The sum of the loading time of the running time and stop by destination component factory First certain track i and l _i, the time required for the total of l _i for each track in consideration of the purpose correcting variable m _i L, constraints When the number of violations is P and the number of tracks is R, the number of tracks can be minimized by directly applying the solution according to the present invention. However, the constraints include the weight and volume that can be loaded on the truck, the working hours of the driver, the combination of parts that can be loaded,
A condition of a time zone during which parts to be delivered can be prepared at each parts factory is added.

FIG. 11 shows an embodiment of the hardware configuration of the system for calculating the solution of the delivery route that minimizes the number of routes according to the present invention. The system 100 includes a central processing unit (CP)
U) 1 and a memory 4. CPU 1 and memory 4
Is connected via a bus 2 to a hard disk device 13 (or MO, CD-ROM 23, DVD
, Etc.) via an IDE controller 25. Similarly, the CPU 1 and the memory 4 are connected via a bus 2 to a hard disk device 30 (or a storage medium drive such as an MO 28, a CD-ROM 23, a DVD, etc.) as an auxiliary storage device via a SCSI controller 27. The floppy disk device 20 is connected to the bus 2 via a floppy disk controller 19.

A floppy disk is inserted into the floppy disk device 20, and the floppy disk or the like or the hard disk device 13 (or MO, CD-ROM,
A storage medium such as a DVD) and a ROM 14 are provided with codes or data of a computer program for giving a command to a CPU or the like in cooperation with an operating system and calculating a solution of a delivery route that minimizes the number of routes of the present invention. Can be recorded and executed by being loaded into the memory 4. The code of the computer program can be compressed or divided into a plurality of pieces and recorded on a plurality of media.

The system 100 further includes user interface hardware and a pointing device (mouse, joystick, etc.) for inputting.
7 or a keyboard 6 and a display 12 for presenting visual data to the user. The solution of the calculated delivery route may be displayed on the display 12, or the solution may be printed on a printer connected via the parallel port 16. The system 100 can connect a modem via the serial port 15. It is possible to communicate with another computer or the like by connecting to a network via the serial port 15 and a modem or a communication adapter 18 (Ethernet or token ring card) or the like. It is also possible to connect a remote transmission / reception device to the serial port 15 or the parallel port 16 to transmit / receive data by infrared rays or radio waves.

The speaker 23 receives, via the amplifier 22, the audio signal that has been D / A (digital / analog converted) by the audio controller 21, and outputs it as audio. Also, the audio controller 21
Converts the audio information received from the microphone 24 into an A / D
(Analog / Digital) conversion, and audio information outside the system can be taken into the system.

As described above, the system of the present invention has various home appliances such as ordinary personal computers (PCs), workstations, notebook PCs, palmtop PCs, network computers, televisions with built-in computers, and communication functions. Game console, telephone, FA
It can be easily understood that the present invention can be implemented by a communication terminal having a communication function including X, a mobile phone, a PHS, an electronic organizer, or the like, or a combination thereof. However, these components are merely examples, and not all of them are essential components of the present invention.

[0029]

According to the present invention, it is possible to generate a delivery route that minimizes the number of routes. The method and system of the present invention can be used without any changes to existing local search methods.
An objective modification method is provided, which is an escape method from a new iterative local search. By using the object correction method of the present invention, it is possible to find an optimal local solution, which is difficult with the conventional iterative local search.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a delivery route.

FIG. 2 is a diagram illustrating an example of a method of expressing a route.

FIG. 3 is a flowchart of a method for obtaining a delivery route.

FIG. 4 is a flowchart of a local search.

FIG. 5 is a flowchart of an evaluation function according to the present invention.

FIG. 6 is a flowchart of updating a best solution.

FIG. 7 is a flowchart of a purpose correction in the present invention.

FIG. 8 is a definition example of an evaluation function according to the present invention.

FIG. 9 is a pseudo code example of a general iterative local search method ILS.

FIG. 10 is an experimental result of evaluating a node sampling method, a route sampling method, and a purpose correction method.

FIG. 11 is an example of a hardware configuration of a system according to the present invention.

FIG. 12 is an example of a delivery route illustrating a solution in which the positions of two points are interchanged.

FIG. 13 is an example of a waypoint arrangement when two points are exchanged.

FIG. 14 is an example of a waypoint arrangement when two consecutive waypoints are exchanged;

Claims (20)

[Claims] 1. A reference solution and a judgment solution are inputted, and a total path length function L, a constraint condition function P, a non-empty path number function R, and
The number of paths is minimized by repeatedly using an evaluation function F for calculating an evaluation result based on a correction flag for correcting a function value for each path for which at least one of the functions L, P, and R is calculated. (1) setting and storing the initial value of the best solution b and the current solution s; and (2) setting the reference solution as the current solution s and the determination solution as the current solution. A step of performing a local search to obtain a local solution t by inputting an evaluation function F as a solution near the solution s and calculating an evaluation value; Inputting the evaluation function F as a local solution t, calculating an evaluation value, and, when the local solution t is a better solution than the best solution b, substituting the local solution t for the best solution b; (4) If the predetermined end condition is satisfied, the process ends. A step that, (5) when the terminating condition is not satisfied substitutes best solutions b the current solution s, by changing the value of the correction flag, the (2) to (4)
Repeating the steps of the above. 2. In the step (2) of obtaining the local solution, a solution included in a set of solutions in which the current solution s is locally changed is set as a solution near the current solution s. 2. The method according to claim 1, wherein a solution obtained by changing the order of the delivery destination is determined as a decision solution. 3. The method according to claim 2, wherein the solution in which the order of the delivery destinations is replaced is a solution in which two or more delivery destinations are replaced, or a solution in which two or more continuous delivery destinations are replaced. . 4. The evaluation function F reduces the value of the constraint condition function P as much as possible, and sets the path number function R
2. A function that evaluates whether the total value of the total path length function L and the value of the non-empty path number function R can be reduced as much as possible and outputs a result.
The described method. 5. An evaluation function F which takes a reference solution s and a decision solution t as inputs and outputs 1 if the decision solution is a better result than the reference solution, and outputs 0 otherwise.
(s, t), (a) If P (t) = 0 and R (s)> R (t), output 1; (b) P (s)> P (t) and L (s) + P (s)> L (t)>
If P (t), output 1; (c) R (s) = R (t) and L (s)
2. The method according to claim 1, wherein the function outputs 1 if + P (s)> L (t)> P (t), and outputs 0 if the condition is not met. 6. The total path length function L according to claim 1, wherein the total path length function L represents a total path length or a total delivery time, and is a total of the path lengths of the respective paths or a total of the delivery times of the respective paths. Method. 7. The non-empty path number function R represents the total number of non-empty paths, and 1 is set if a certain path is not empty, and 0 if it is empty. The method of claim 1, wherein 8. The constraint function P represents the total value of the number of condition violations, and is the sum of the case where the maximum route length or the maximum delivery time is exceeded and the sum of the number of deliveries exceeding the load capacity.
The method of claim 1. 9. The step (5) of changing the correction flag includes:
If the termination condition is not satisfied, the best solution b is substituted for the current solution s, and (a) the value of the correction flag is set so that all the routes are not affected by the correction flag, and (b)
Choose a non-empty route at random, set the value of the correction flag so that each is affected by the correction flag,
2. The method according to claim 1, further comprising the step of: (c) setting the value of the correction flag so as to be affected by the correction flag for all of the empty routes, and repeating the steps (2) to (4). 10. A reference solution and a judgment solution are input, and a total path length function L, a constraint condition function P, a non-empty path number function R, and an object of calculation of at least one of the functions L, P, R A correction flag for correcting a function value for each route, and a solution for a delivery route that minimizes the number of routes by repeatedly using an evaluation function F for calculating an evaluation result. (1) an initial value setting means for setting and storing the best solution b and the current solution s as initial values, and (2) a reference solution as a current solution s and a judgment solution as a solution near the current solution s. , A local solution calculating means for performing a local search to obtain a local solution t by calculating an evaluation value, and (3) a best solution b as a reference solution, and a local solution t as a decision solution. Is input to the evaluation function F, and the evaluation value is calculated. a best solution calculating means for substituting the local solution t for the best solution b when t is a better solution than the best solution b; and (4) substituting the best solution b for the current solution s, Means for changing the value of. 11. A local solution calculating means (2) for obtaining said local solution, wherein said local solution is included in a set of locally modified solutions when said determination solution is a solution near said current solution s. The system according to claim 10, wherein a solution obtained by changing the order of the delivery destination is determined as a decision solution. 12. The solution in which the order of the delivery destination is changed is 2
12. The system of claim 11, wherein the solution is to swap one or more destinations or to swap two or more consecutive destinations. 13. The evaluation function F, wherein the value of the constraint function P is reduced as much as possible, the value of the non-empty path number function R is reduced as much as possible, and the total path length function L and the value of the non-empty path number function R are reduced as much as possible. The system according to claim 10, wherein the function is a function that evaluates whether or not the sum can be reduced and outputs a result. 14. The evaluation function F includes a reference solution s and a judgment solution t.
The evaluation function F that outputs 1 if the decision solution is a better result than the reference solution and outputs 0 otherwise
(s, t), (a) If P (t) = 0 and R (s)> R (t), output 1; (b) P (s)> P (t) and L (s) + P (s)> L (t)>
If P (t), output 1; (c) R (s) = R (t) and L (s)
The system according to claim 10, wherein the function outputs 1 if + P (s)> L (t)> P (t), and outputs 0 if the above condition is not met. 15. The total path length function L represents a total path length or a total delivery time, and is a sum of path lengths of respective paths,
The system according to claim 10, wherein the total is the delivery time of each route. 16. The non-empty path number function R represents the total number of non-empty paths, and 1 is set when a certain path is not empty, and 0 when it is empty, and the values are summed over all paths. The system of claim 10, wherein 17. The method according to claim 17, wherein said constraint function P represents a total value of the number of condition violations, and is a sum of a case where the maximum route length or the maximum delivery time is exceeded and a sum of the number of deliveries exceeding the load capacity. Item 11. The system according to Item 10. 18. A system for calculating a solution of a delivery route for minimizing the number of routes for a cash transport vehicle for exchanging ATM cassettes, wherein the total route length function L is calculated based on the travel time of the cash transport vehicle. And the cassette exchange time of the ATM, and the constraint function P is the number of violations of constraints including the number of cassettes that can be loaded on the cash transport vehicle or the working hours of the driver. 11. The system of claim 10, wherein the number is a cash transport vehicle. 19. A system for calculating a solution of a delivery route for minimizing the number of routes for a truck that travels around a parts factory and transports parts to an assembly factory, wherein the total path length function L is a truck. Is the sum of the travel time of the vehicle and the loading time at the drop-off parts factory. The constraint function P is the weight and volume that can be loaded on the truck, the working hours of the driver,
The system according to claim 10, wherein the number of cases is a combination of parts that can be mounted together or the number of violations of a constraint condition including a time zone in which parts can be prepared at each parts factory, and the non-empty path number function R is the number of trucks. . 20. A reference solution and a judgment solution are input, and a total path length function L, a constraint condition function P, a non-empty path number function R, and a calculation object of at least one of the functions L, P, and R A medium including a program for calculating a solution of a delivery route that minimizes the number of routes by repeatedly using an evaluation function F for calculating an evaluation result based on a correction flag for correcting a function value for each route to be obtained The program has (1) a function of setting and storing the best solution b and the current solution s as initial values, and (2) a function of setting a reference solution as the current solution s and a method of determining a judgment solution near the current solution s. As a solution, a function of obtaining a local solution t by performing a local search by inputting it to an evaluation function F and calculating an evaluation value; Input to the evaluation function F, calculate the evaluation value, and But when was the solution better than best solution b, a function of substituting the local solution t the best solution b,
(4) A function for terminating the process when a predetermined end condition is satisfied, and (5) When the end condition is not satisfied, the best solution b is substituted for the current solution s, and the correction flag is set. A medium including a program, having a function of changing a value and repeating the functions (2) to (4).