CN113762780A - Method for processing medical waste collection problem by using improved genetic algorithm - Google Patents

Abstract

The invention discloses a method for using an improved genetic algorithm to deal with the problem of medical waste collection. Multiple robots are used to jointly complete the medical waste collection work, and the scheduling scheme of the multiple robots is obtained through the following steps: S1. According to the plan of the hospital, determine the map Calculate the distance between the collection point and the workstation and the distance between the collection point and the collection point, get the distance matrix, and determine the amount of medical waste at each collection point; S2 establishes the collection point for medical waste. Robot scheduling model; S3, solving the robot scheduling model for medical waste collection established in step S2 through an improved genetic algorithm, thereby obtaining a scheduling scheme of multiple robots. The invention has the advantages of improving the efficiency of medical waste collection, improving the safety and usability in the collection process, avoiding the algorithm falling into the "extreme value trap" and the like.

Description

A method for dealing with medical waste collection problems using an improved genetic algorithm

technical field

The invention relates to the technical fields of computers, medical waste collection and robot path planning, and in particular to a method for dealing with the problem of medical waste collection by using an improved genetic algorithm.

Background technique

With the continuous development of artificial intelligence technology, more and more jobs can be done by robots. For large medical institutions, a large amount of infectious health care waste (IHCW) is generated every day. Using robots to collect medical waste can not only improve the processing capacity, but also avoid the risk of infection of the processing personnel. Inside the hospital, after the robot workstation is set up, the amount of infectious medical waste generated by each collection point is different. It is necessary to find a driving path from the workstation to each scattered collection point and meet certain constraints. The problem can be transformed into a classic Vehicle Routing Problem (VRP).

The logistics transportation scheduling problem has always been a hot research problem in logistics distribution, and it is widely used in fields such as transportation, industrial management, and logistics transportation. Scholars at home and abroad have carried out research on it from the directions of path construction, local search, and mathematical programming. To sum up, the most effective methods to solve the logistics distribution path optimization problem are: ant colony algorithm, tabu search algorithm, simulated annealing algorithm, genetic algorithm. The ant colony algorithm itself is very complex, requires a long search time, and is prone to stagnation. The tabu search algorithm is a single operation, and there can only be one initial solution in the search process, and it has a strong dependence on the initial solution. Simulated annealing algorithm has slow convergence speed and long execution time, and its performance has a great relationship with the initial value.

SUMMARY OF THE INVENTION

The purpose of the present invention is to overcome the deficiencies of the prior art, and to provide a method for dealing with the problem of medical waste collection using an improved genetic algorithm, which not only improves the efficiency of medical waste collection, the safety and usability during the collection process, but also avoids the algorithm falling into "Extreme trap" and other advantages.

For achieving the above object, the technical scheme provided by the present invention is:

A method for using an improved genetic algorithm to deal with the problem of medical waste collection. Multiple robots are used to jointly complete the medical waste collection work, and the scheduling scheme of the multiple robots is obtained through the following steps:

S1. According to the plan of the hospital, determine the location of each collection point and robot workstation in the figure, calculate the distance between the collection point and the workstation and the distance between the collection point and the collection point, obtain a distance matrix, and determine the amount of medical waste at each collection point ;

S2 establishes a robot scheduling model for medical waste collection;

S3. Solve the robot scheduling model for medical waste collection established in step S2 through an improved genetic algorithm, thereby obtaining scheduling schemes for multiple robots.

Further, in the step S2, the established robot scheduling model for medical waste collection is a robot scheduling model considering the upper semi-soft time window, and the establishment process is as follows:

S2-1. Definition of semi-soft time window, the robot can reach outside the time window, including the following situations:

When the robot reaches the collection point within the specified time window [a _i , b _i ], it will not be penalized;

If the robot arrives earlier than a _i , it will not be penalized, thus shortening its stay in the bin, but if the robot arrives later than b _i , it will be penalized;

Let the allowable arrival time be b′ _i (b′ _i >b _i ):

b′ _i =b ₀ -t _i0 (1)

In formula (1), b ₀ is a constant, representing the time when the robot finally returns to the workstation; t _i0 represents the travel time from the collection point i to the workstation; the formula stipulates that enough time should be reserved from the collection point i to the workstation;

S2-2. Definition of timeout penalty, when the robot arrives at the time interval [b _i , b′ _i ], that is, the service start time s _i falls in [b _i , b′ _i ], it will be penalized c _delay ×(s _i -b _i );

The penalty for each collection point i is calculated as follows:

In formula (2), s _i is the service time starting from collection point i, and c _delay is the unit cost of delay time;

S2-3. Define the total travel time M of each route for the time limit:

为布尔变量，若机器人k在时间轮h中直接从收集点i移动到收集点j，则该值为1，否则，该值为0；t_ij为从收集点i移动到收集点j的移动时间；in,

is a Boolean variable, if the robot k moves directly from the collection point i to the collection point j in the time wheel h, the value is 1, otherwise, the value is 0; t _ij is the movement from the collection point i to the collection point j time;

S2-4. Set the objective function to minimize the total cost, including the startup cost, travel distance cost and delay time cost of the robot:

为布尔变量，若机器人k在时间轮h中直接从工作站0移动到收集点i，则该值为1；否则，该值为0；c_start表示机器人的启动成本；c_ij为从收集点i到收集点j的移动距离；

为布尔变量，若收集点i在时间轮h中由机器人k服务，则值为1，否则，该值为0；In formula (3),

is a Boolean variable, if the robot k moves directly from the workstation 0 to the collection point i in the time wheel h, the value is 1; otherwise, the value is 0; c _start represents the start-up cost of the robot; c _ij is from the collection point i The moving distance to the collection point j;

is a Boolean variable, if the collection point i is served by the robot k in the time wheel h, the value is 1, otherwise, the value is 0;

Restrictions:

(T _i -1)M≤s _i -b _i ≤T _i M (13)

Equations (4) and (5) are degree constraints, constraint (4) ensures that each acquisition point can only be accessed once; constraint (5) ensures that the number of incoming arcs is equal to the number of nodes leaving arcs; constraint (6) ensures that each acquisition Points can only be assigned to one robot; Constraint (7) indicates that the distance should be a non-negative variable; Constraint (8) ensures that the total volume on each route does not exceed the robot's capacity, where the amount of medical waste at the ith medical waste collection point , Q represents the capacity of the robot to store medical waste; constraint (9) calculates the service start time of each collection point in the route; constraint (10) limits the service start time between its earliest and latest arrival times; constraint ( 11) Make the service start time less than the total time of each route; Constraint (12) indicates that the service time should be a non-negative variable; Constraint (13) determines whether the service start time of the collection point is within its corresponding penalty interval, i.e. (b _i , b′ _i ]; constraints (14)-(16) are integrity constraints.

Further, the specific process of the step S3 is as follows:

S3-1. Input parameters: medical waste collection requirements, travel distance, travel time, time window;

S3-2. According to the number of robots and the initialization solution of the medical waste collection point, multiple paths are generated, and the number of iterations is set, k=0;

S3-3. Calculate the fitness of each current path;

S3-4, select an initial path with high fitness according to fitness;

S3-5, the initial path obtained in step S3-4 is divided into two groups of nodes, the first group is placed in the front, and the start point and the end point of the first group are both 0;

S3-6, the second group of nodes is the first group of unvisited nodes, and the nodes of this group are arranged in order;

S3-7. By inserting 0 multiple times, the second group of nodes is randomly divided into 1 to n-1 paths, and the objective function of each path is calculated;

S3-8. Sort the paths according to the objective function from small to large to form a list;

S3-9. Select the route with the lowest total cost, that is, the smallest objective function, and judge whether it is feasible or not, and delete it directly if it is not feasible; the feasible route is attached to the back of the first group of routes to form a brand-new route, and a candidate solution is obtained; ( Steps S3-5 to S3-9 can refer to the example given in FIG. 3)

S3-10, the mutation operator flips the bit to the complementary bit;

S3-11, determine whether the current candidate solution is feasible, if feasible, calculate its objective function, otherwise return to step S3-10;

S3-12. Select the candidate solution with the lowest total cost, that is, the smallest objective function;

S3-13, determine whether k is greater than the set number of iterations, if so, the candidate solution obtained in step S3-12 is the final optimal solution, otherwise, return to step S3-3.

Further, in the step S3-3, the calculated fitness is used to evaluate the current solution, and the calculation formula is as follows:

In formula (17), f is the fitness value, and Z is the total cost.

Compared with the prior art, the principle and advantages of this scheme are as follows:

1. The medical waste collection model of medical institutions is represented by an improved vehicle scheduling problem with a semi-soft time window, and the constraints on the modeling of the problems that need to be paid attention to when dealing with medical wastes improve the safety and usability in the collection process. .

2. The improved genetic algorithm is used to solve the robot scheduling model, and the "sequential crossover" combined with "crossover mutation" strategy is used to further optimize the algorithm, which effectively increases the population diversity of the algorithm and avoids the algorithm from falling into the "extreme trap".

3. Use multiple robots to work together to improve the efficiency of medical waste collection.

Description of drawings

In order to illustrate the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the services required in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only For some embodiments of the present invention, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

Fig. 1 is a kind of principle flow chart of problem solving in a method of using an improved genetic algorithm to deal with the medical waste collection problem of the present invention;

2 is a schematic diagram of a medical waste collection environment in an embodiment of the present invention;

Figure 3 is a schematic diagram of the genetic algorithm sequence crossover and feasibility detection principle;

4 is a schematic diagram of a distance matrix involved in an embodiment of the present invention;

5 is a statistical diagram of the distribution of medical wastes at different collection points in the embodiment of the present invention;

6 is an output result diagram of an improved genetic algorithm in an embodiment of the present invention;

FIG. 7 is an optimization effect diagram of an improved genetic algorithm in an embodiment of the present invention.

Detailed ways

Below in conjunction with specific embodiment, the present invention will be further described:

In a method for using an improved genetic algorithm to deal with the problem of medical waste collection described in this embodiment, a plurality of robots are used to jointly complete the medical waste collection work, and the scheduling scheme of the plurality of robots is obtained through the following steps:

S1. According to the plan view of the hospital as shown in Figure 2, determine the positions of the collection points and the robot workstation in the figure, calculate the distance between the collection point and the workstation and the distance between the collection point and the collection point, and obtain as shown in Figure 4 A distance matrix to determine the amount of medical waste at each collection point as shown in Figure 5;

S2. The established robot scheduling model for medical waste collection is a robot scheduling model considering the upper semi-soft time window. The establishment process is as follows:

S2-1. Definition of semi-soft time window, the robot can reach outside the time window, including the following situations:

When the robot reaches the collection point within the specified time window [a _i , b _i ], it will not be penalized;

If the robot arrives earlier than a _i , it will not be penalized, thus shortening its stay in the bin, but if the robot arrives later than b _i , it will be penalized;

Let the allowable arrival time be b′ _i (b′ _i >b _i ):

b′ _i =b ₀ -t _i0 (1)

In formula (1), b ₀ is a constant, representing the time when the robot finally returns to the workstation; t _i0 represents the travel time from the collection point i to the workstation; the formula stipulates that enough time should be reserved from the collection point i to the workstation;

S2-2. Definition of timeout penalty, when the robot arrives at the time interval [b _i , b′ _i ], that is, the service start time s _i falls in [b _i , b′ _i ], it will be penalized c _delay ×(s _i -b _i );

The penalty for each collection point i is calculated as follows:

In formula (2), s _i is the service time starting from collection point i, and c _delay is the unit cost of delay time;

S2-3. Define the total travel time M of each route for the time limit:

为布尔变量，若机器人k在时间轮h中直接从收集点i移动到收集点j，则该值为1，否则，该值为0；t_ij为从收集点i移动到收集点j的移动时间；in,

is a Boolean variable, if the robot k moves directly from the collection point i to the collection point j in the time wheel h, the value is 1, otherwise, the value is 0; t _ij is the movement from the collection point i to the collection point j time;

S2-4. Set the objective function to minimize the total cost, including the startup cost, travel distance cost and delay time cost of the robot:

为布尔变量，若机器人k在时间轮h中直接从工作站0移动到收集点i，则该值为1；否则，该值为0；c_start表示机器人的启动成本；c_ij为从收集点i到收集点j的移动距离；

为布尔变量，若收集点i在时间轮h中由机器人k服务，则值为1，否则，该值为0；In formula (3),

is a Boolean variable, if the robot k moves directly from the workstation 0 to the collection point i in the time wheel h, the value is 1; otherwise, the value is 0; c _start represents the start-up cost of the robot; c _ij is from the collection point i The moving distance to the collection point j;

is a Boolean variable, if the collection point i is served by the robot k in the time wheel h, the value is 1, otherwise, the value is 0;

Restrictions:

(T _i -1)M≤s _i -b _i ≤T _i M (13)

Equations (4) and (5) are degree constraints, constraint (4) ensures that each acquisition point can only be accessed once; constraint (5) ensures that the number of incoming arcs is equal to the number of nodes leaving arcs; constraint (6) ensures that each acquisition Points can only be assigned to one robot; Constraint (7) indicates that the distance should be a non-negative variable; Constraint (8) ensures that the total volume on each route does not exceed the robot's capacity, where the amount of medical waste at the ith medical waste collection point , Q represents the capacity of the robot to store medical waste; constraint (9) calculates the service start time of each collection point in the route; constraint (10) limits the service start time between its earliest and latest arrival times; constraint ( 11) Make the service start time less than the total time of each route; Constraint (12) indicates that the service time should be a non-negative variable; Constraint (13) determines whether the service start time of the collection point is within its corresponding penalty interval, i.e. (b _i , b _i ]; constraints (14)-(16) are integrity constraints.

S3. Solve the robot scheduling model for medical waste collection established in step S2 through an improved genetic algorithm, thereby obtaining scheduling schemes for multiple robots.

As shown in Figure 1, the specific process is as follows:

S3-1. Input parameters: medical waste collection requirements, travel distance, travel time, time window;

S3-2. According to the number of robots and the initialization solution of the medical waste collection point, multiple paths are generated, and the number of iterations is set, k=0;

S3-3. Calculate the fitness of each current path;

The calculated fitness is used to evaluate the current solution and is calculated as:

In formula (17), f is the fitness value, and Z is the total cost.

S3-4, select an initial path with high fitness according to fitness;

S3-5, the initial path obtained in step S3-4 is divided into two groups of nodes, the first group is placed in the front, and the start point and the end point of the first group are both 0;

S3-6, the second group of nodes is the first group of unvisited nodes, and the nodes of this group are arranged in order;

S3-7. By inserting 0 multiple times, the second group of nodes is randomly divided into 1 to n-1 paths, and the objective function of each path is calculated;

S3-8. Sort the paths according to the objective function from small to large to form a list;

S3-9. Select the route with the lowest total cost, that is, the smallest objective function, and judge whether it is feasible or not, and delete it if it is not feasible; the feasible route is appended to the back of the first group of routes to form a new route, and a candidate solution is obtained;

S3-10, the mutation operator flips the bit to the complementary bit;

S3-11, determine whether the current candidate solution is feasible, if feasible, calculate its objective function, otherwise return to step S3-10;

S3-12. Select the candidate solution with the lowest total cost, that is, the smallest objective function;

S3-13, determine whether k is greater than the set number of iterations, if so, the candidate solution obtained in step S3-12 is the final optimal solution, otherwise, return to step S3-3.

As shown in Figure 6, the final output is as follows:

Route for robot 1: 0-2-6-1-11-19-22-25-21-27-0;

Route for robot 2: 0-7-10-13-9-15-5-0;

Route for robot 3: 0-26-23-24-14-8-20-17-0;

Route for Robot 4: 0-12-3-4-28-18-16-0.

The optimization effect is shown in Figure 7.

The above-mentioned embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of implementation of the present invention. Therefore, any changes made according to the shape and principle of the present invention should be included within the protection scope of the present invention.

Claims (4)

1. a method for using an improved genetic algorithm to deal with the problem of medical waste collection, is characterized in that, by a plurality of robots to jointly complete the medical waste collection work, and the scheduling scheme of a plurality of robots is obtained by the following steps: S1. According to the plan of the hospital, determine the location of each collection point and robot workstation in the figure, calculate the distance between the collection point and the workstation and the distance between the collection point and the collection point, obtain a distance matrix, and determine the amount of medical waste at each collection point ; S2. Establish a robot scheduling model for medical waste collection; S3. Solve the robot scheduling model for medical waste collection established in step S2 through an improved genetic algorithm, thereby obtaining scheduling schemes for multiple robots. 2. a kind of method that uses improved genetic algorithm to deal with the problem of medical waste collection according to claim 1, is characterized in that, in described step S2, the robot scheduling model that is established for medical waste collection is to consider upper half soft time The robot scheduling model of the window is established as follows: S2-1. Definition of semi-soft time window, the robot can reach outside the time window, including the following situations: When the robot reaches the collection point within the specified time window [a _i , b _i ], it will not be penalized; If the robot arrives earlier than a _i , it will not be penalized, thus shortening its stay in the bin, but if the robot arrives later than b _i , it will be penalized; Let the allowable arrival time be b′ _i (b′ _i >b _i ): b′ _i =b ₀ -t _i0 (1) In formula (1), b ₀ is a constant, representing the time when the robot finally returns to the workstation; t _i0 represents the travel time from the collection point i to the workstation; the formula stipulates that enough time should be reserved from the collection point i to the workstation; S2-2. Definition of timeout penalty, when the robot arrives at the time interval [b _i , b′ _i ], that is, the service start time s _i falls in [b _i , b′ _i ], it will be penalized c _delay ×(s _i -b _i ); The penalty for each collection point i is calculated as follows:

In formula (2), s _i is the service time starting from collection point i, and c _delay is the unit cost of delay time; S2-3. Define the total travel time M of each route for the time limit:

in,

is a Boolean variable, if the robot k moves directly from the collection point i to the collection point j in the time wheel h, the value is 1, otherwise, the value is 0; t _ij is the movement from the collection point i to the collection point j time; S2-4. Set the objective function to minimize the total cost, including the startup cost, travel distance cost and delay time cost of the robot:

In formula (3),

is a Boolean variable, if the robot k moves directly from the workstation 0 to the collection point i in the time wheel h, the value is 1; otherwise, the value is 0; c _start represents the start-up cost of the robot; c _ij is from the collection point i The moving distance to the collection point j;

is a Boolean variable, if the collection point i is served by the robot k in the time wheel h, the value is 1, otherwise, the value is 0; Restrictions:

(T _i -1)M≤s _i -b _i ≤T _i M (13)

Equations (4) and (5) are degree constraints, constraint (4) ensures that each acquisition point can only be accessed once; constraint (5) ensures that the number of incoming arcs is equal to the number of nodes leaving arcs; constraint (6) ensures that each acquisition Points can only be assigned to one robot; Constraint (7) indicates that the distance should be a non-negative variable; Constraint (8) ensures that the total volume on each route does not exceed the robot's capacity, where the amount of medical waste at the ith medical waste collection point , Q represents the capacity of the robot to store medical waste; constraint (9) calculates the service start time of each collection point in the route; constraint (10) limits the service start time between its earliest and latest arrival times; constraint ( 11) Make the service start time less than the total time of each route; Constraint (12) indicates that the service time should be a non-negative variable; Constraint (13) determines whether the service start time of the collection point is within its corresponding penalty interval, i.e. (b _i , b′ _i ]; constraints (14)-(16) are integrity constraints. 3. a kind of method that uses improved genetic algorithm to handle medical waste collection problem according to claim 2, is characterized in that, the concrete process of described step S3 is as follows: S3-1. Input parameters: medical waste collection requirements, travel distance, travel time, time window; S3-2. According to the number of robots and the initialization solution of the medical waste collection point, multiple paths are generated, and the number of iterations is set, k=0; S3-3. Calculate the fitness of each current path; S3-4, select an initial path with high fitness according to fitness; S3-5, the initial path obtained in step S3-4 is divided into two groups of nodes, the first group is placed in the front, and the start point and the end point of the first group are both 0; S3-6, the second group of nodes is the first group of unvisited nodes, and the nodes of this group are arranged in order; S3-7. By inserting 0 multiple times, the second group of nodes is randomly divided into 1 to n-1 paths, and the objective function of each path is calculated; S3-8. Sort the paths according to the objective function from small to large to form a list; S3-9. Select the route with the lowest total cost, that is, the smallest objective function, and judge whether it is feasible or not, and delete it if it is not feasible; the feasible route is appended to the back of the first group of routes to form a new route, and a candidate solution is obtained; S3-10, the mutation operator flips the bit to the complementary bit; S3-11, determine whether the current candidate solution is feasible, if feasible, calculate its objective function, otherwise return to step S3-10; S3-12. Select the candidate solution with the lowest total cost, that is, the smallest objective function; S3-13, determine whether k is greater than the set number of iterations, if so, the candidate solution obtained in step S3-12 is the final optimal solution, otherwise, return to step S3-3. 4. A method of using an improved genetic algorithm to deal with the problem of medical waste collection according to claim 3, wherein in the step S3-3, the calculated fitness is used to evaluate the current solution, and the calculation formula is as follows :

In formula (17), f is the fitness value, and Z is the total cost.

Images