Route planning method for hospital disinfection robots

Abstract

A composite brainstorming optimization method for hospital disinfection robot route planning solves the problem of hospital disinfection robot route planning. Aiming at the disadvantage of being easily trapped in local optimality, the idea of ​​simulated annealing and Metropolis function improvement are integrated. Since the idea of ​​simulated annealing is that there is a certain probability of accepting non-optimal solutions as new solutions under the Metropolis criterion, the selection of new solutions is adjusted in the brainstorming optimization method of the present invention, so that the algorithm is not easily trapped in local optimality.

Description

Technical Field

[0001] The invention belongs to the technical field of intelligent mobile robots and is a route planning method for a hospital disinfection robot. Background Art

[0002] With the continuous improvement of artificial intelligence technology and modern medical service levels, intelligent disinfection robots have become an important equipment for future hospital infection prevention and control, providing a new working mode for the disinfection of traditional medical spaces. It can not only reduce the heavy workload of manual work, but also improve the efficiency of disinfection operations. However, due to the particularity of hospital environment, public facilities, and medical staff, the safety of disinfection robot operations is facing many challenges.

[0003] Hospital disinfection robot route planning refers to finding a collision-free optimal route from the starting point through all disinfection points in a complex environment. According to the characteristics of the hospital disinfection robot route planning, the hospital disinfection robot route planning problem can be transformed into a Traveling Saleman Problem (TSP) problem. The Traveling Salesman Problem (TSP) is a typical NP-complete problem, that is, its worst-case time complexity increases exponentially with the increase of the problem scale, and so far, an effective polynomial time algorithm has not been found. At present, the commonly used algorithms for studying the classic TSP include the firefly algorithm, the simulated annealing algorithm, and the neural network algorithm. The firefly algorithm is easy to fall into the local optimal solution; the simulated annealing algorithm is simple to operate and may jump out of the local optimal solution in the later stage of the algorithm, but it has the disadvantages of slow convergence speed, long execution time, algorithm performance is related to the initial value, and parameter sensitivity; the simulated annealing algorithm has the disadvantages of slow convergence speed, long execution time, algorithm performance is related to the initial value, and parameter sensitivity; the neural network algorithm is difficult to implement in a more complex environment.

[0004] Brainstorm optimization (BSO) is a new swarm intelligence algorithm based on the idea of ​​human creative problem solving. Since the BSO algorithm has no requirements for initial values, it has strong robustness and fast convergence ability, and has very strong global search ability. The brainstorm optimization method is applied to the study of hospital disinfection robot route planning problems, which can effectively solve such problems. However, the algorithm still has the problem of fast convergence speed in the early stage of iteration and easy to fall into local optimality too early. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a composite brainstorming optimization method to solve the route planning of hospital disinfection robots. Aiming at the disadvantage of being easily trapped in local optimality, the idea of ​​simulated annealing and improvement of Metropolis function are incorporated. Since the idea of ​​simulated annealing is that there is a certain probability of accepting non-optimal solutions as new solutions under the Metropolis criterion, the selection of new solutions is adjusted in the brainstorming optimization method of the present invention, so that the algorithm is not easily trapped in local optimality.

[0006] The technical solution adopted by the present invention to achieve the above-mentioned invention object is as follows:

[0007] A route planning method for a hospital disinfection robot includes the following steps:

[0008] Step 1: Use the camera, laser radar, and infrared sensor of the disinfection robot to collect the working environment information of the disinfection robot and the coordinate information of all disinfection points, and rasterize the map;

[0009] Step 2: Using the coordinate information of each disinfection point, calculate the shortest mutual obstacle avoidance distance between disinfection points and establish a distance matrix;

[0010] Step 3: Encode all disinfection points with integers; a route represents a potential feasible solution. In disinfection robot route planning, a route represents the sequence of disinfection points that the robot passes through;

[0011] Step 4: Set the initial parameters, including the total number of routes n, the number of disinfection points m, the maximum number of groups K, the probability value of selecting the number of central routes p, the initial value of the number of iterations iter = 1, the starting temperature Ts = 500, the cooling ratio η = 0.99, and the ending temperature Tz = 0.1;

[0012] Step 5: Generate n routes randomly to form a route population, and calculate the fitness value of each route using formula (1):

[0013]

[0014] Among them, D(v i-1 ,v i ) indicates the disinfecting point v i-1 to v i The shortest obstacle avoidance distance, v i Represents the position of the i-th disinfection point;

[0015] Step 6: k-means clustering, the process is as follows:

[0016] a) Randomly select K center routes μ k The initial value of , k∈[1,K];

[0017] b) Classify each route into the class represented by the nearest central route according to its fitness value;

[0018] c) Update the center route μ k The value of is the mean of each category;

[0019] d) Repeat (b) and (c) until the maximum number of iterations is reached and output the classification result;

[0020] Step 7: Randomly generate a random number R∈(0,1). If the random number is less than p, go to step 8, otherwise go to step 9;

[0021] Step 8: Single route optimization, randomly select a route and perform insertion operation to generate a new route;

[0022] Step 9: Optimization between two routes, randomly selecting routes from two different groups for crossover operation to generate a new route;

[0023] Step 10: Probability through Monte Carlo Function Update the route population and obtain new routes;

[0024] Where Tn is the current temperature, E(X new ) is the fitness value of the new route, E(X old ) The fitness value of the old route;

[0025] Step 11: Generate a random number R∈(0,1). If R>p, keep the newly generated route, otherwise do not keep it.

[0026] Step 12: Update the optimal route of the route population;

[0027] Step 13: Have all routes been updated? If yes, jump to step 14; otherwise, jump to step 7;

[0028] Step 14: Whether the termination temperature Tz is reached, if yes, the optimal route is output, otherwise, the cooling operation is performed and the process goes to step 6;

[0029] Step 15: Based on the route obtained in step 14, use the A* algorithm to plan a path in the map.

[0030] The beneficial effects of the present invention are mainly manifested in: since the traditional brainstorming optimization method is prone to fall into the local optimum too early, in order to integrate the idea of ​​simulated annealing and the improvement of Metropolis function, since the idea of ​​simulated annealing is that there is a certain probability of accepting non-optimal solutions as new solutions under the Metropolis criterion, the diversity of the algorithm is improved, thereby increasing the probability of the algorithm jumping out of the local optimum. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flow chart of the present invention.

[0032] Figure 2 This is the route planning result of the traditional brainstorming algorithm, where the white area represents the barrier-free drivable area, the black area represents the obstacle, the dark gray area represents the disinfection point, and the light gray area is the actual driving map of the robot.

[0033] Figure 3 This is the route planning result of the composite brainstorming algorithm, where the white area represents the barrier-free drivable area, the black area represents the obstacle, the dark gray area represents the disinfection point, and the light gray area represents the actual driving map of the robot. DETAILED DESCRIPTION

[0034] The present invention will be further described below in conjunction with the accompanying drawings.

[0035] Reference Figure 1 to Figure 3 , a route planning method for a hospital disinfection robot, comprising the following steps:

[0036] Step 1: Use the camera, laser radar, and infrared sensor of the disinfection robot to collect the working environment information of the disinfection robot and the coordinate information of all disinfection points, and rasterize the map;

[0037] Step 2: Using the coordinate information of each disinfection point, calculate the shortest mutual obstacle avoidance distance between disinfection points and establish a distance matrix;

[0038] Step 3: Encode all disinfection points with integers; a route represents a potential feasible solution. In disinfection robot route planning, a route represents the sequence of disinfection points that the robot passes through;

[0039] Step 4: Set the initial parameters, including the total number of routes n, the number of disinfection points m, the maximum number of groups K, the probability value of selecting the number of central routes p, the initial value of the number of iterations iter = 1, the starting temperature Ts = 500, the cooling ratio η = 0.99, and the ending temperature Tz = 0.1;

[0040] Step 5: Generate n routes randomly to form a route population, and calculate the fitness value of each route using formula (1):

[0041]

[0042] Among them, D(v i-1 ,v i ) indicates the disinfecting point v i-1 to v i The shortest obstacle avoidance distance, v i Represents the position of the i-th disinfection point;

[0043] Step 6: k-means clustering, the process is as follows:

[0044] a) Randomly select K center routes μ k The initial value of , k∈[1,K];

[0045] b) Classify each route into the class represented by the nearest central route according to its fitness value;

[0046] c) Update the center route μ k The value of is the mean of each category;

[0047] d) Repeat (b) and (c) until the maximum number of iterations is reached and output the classification result;

[0048] Step 7: Randomly generate a random number R∈(0,1). If the random number is less than p, go to step 8, otherwise go to step 9;

[0049] Step 8: Single route optimization, randomly select a route and perform insertion operation to generate a new route;

[0050] Step 9: Optimization between two routes, randomly selecting routes from two different groups for crossover operation to generate a new route;

[0051] Step 10: Probability through Monte Carlo Function Update the route population and obtain new routes;

[0052] Where Tn is the current temperature, E(X new ) is the fitness value of the new route, E(X old ) The fitness value of the old route;

[0053] Step 11: Generate a random number R∈(0,1). If R>p, keep the newly generated route, otherwise do not keep it.

[0054] Step 12: Update the optimal route of the route population;

[0055] Step 13: Have all routes been updated? If yes, jump to step 14; otherwise, jump to step 7;

[0056] Step 14: Whether the termination temperature Tz is reached, if yes, the optimal route is output, otherwise, the cooling operation is performed and the process goes to step 6;

[0057] Step 15: Based on the route obtained in step 14, use the A* algorithm to plan a path in the map.

[0058] In order to verify the feasibility and effectiveness of the present invention, simulation experiments were conducted using maps of the same size and the same number of disinfection points. The computer performance parameters used for simulation were: AMD Ryzen 54600U with Radeon Graphics 2.10GHz, 16GB of memory running on Window 11, and Matlab 2020b software was used to conduct simulation experiments, verifying the route planning of 39 disinfection points on an 80×170 grid map.

[0059] Under the same map environment and the same number of disinfection points, simulation experiments were carried out using the traditional method and the method proposed in this paper. The experimental results are as follows: Figure 2 and Figure 3 As shown;

[0060] The experimental results show that under the same conditions, the total length of the route planned by the brainstorming optimization method with the improved simulation algorithm is shorter. Figure 3 As shown, the method of the present invention is superior to the route planning method of the hospital disinfection robot.

[0061] The above is an embodiment of the present invention. It should be pointed out that for those skilled in the art, any modification or adaptation thereof without departing from the present invention is deemed to be within the protection scope of the present invention.

Claims

1. A route planning method for a hospital disinfection robot, characterized in that: The method comprises the following steps: Step 1: Use the camera, laser radar, and infrared sensor of the disinfection robot to collect the working environment information of the disinfection robot and the coordinate information of all disinfection points, and rasterize the map; Step 2: Using the coordinate information of each disinfection point, calculate the shortest mutual obstacle avoidance distance between disinfection points and establish a distance matrix; Step 3: Encode all disinfection points with integers; a route represents a potential feasible solution. In disinfection robot route planning, a route represents the sequence of disinfection points that the robot passes through; Step 4: Set the initial parameters, including the total number of routes n, the number of disinfection points m, the maximum number of groups K, the probability value of selecting the number of central routes p, the initial value of the number of iterations iter = 1, the starting temperature Ts = 500, the cooling ratio η = 0.99, and the ending temperature Tz = 0.1; Step 5: Generate n routes randomly to form a route population, and calculate the fitness value of each route using formula (1): Among them, D(v i-1 ,v i ) indicates the disinfecting point v i-1 to v i The shortest obstacle avoidance distance, v i Represents the position of the i-th disinfection point; Step 6: k-means clustering, the process is as follows: a) Randomly select K center routes μ k The initial value of , k∈[1,K]; b) Classify each route into the class represented by the nearest central route according to its fitness value; c) Update the center route μ k The value of is the mean of each category; d) Repeat (b) and (c) until the maximum number of iterations is reached and output the classification result; Step 7: Randomly generate a random number R∈(0,1). If the random number is less than p, go to step 8, otherwise go to step 9; Step 8: Single route optimization, randomly select a route and perform insertion operation to generate a new route; Step 9: Optimization between two routes, randomly selecting routes from two different groups for crossover operation to generate a new route; Step 10: Probability through Monte Carlo Function Update the route population and obtain new routes; Where Tn is the current temperature, E(X new ) is the fitness value of the new route, E(X old ) The fitness value of the old route; Step 11: Generate a random number R∈(0,1). If R>p, keep the newly generated route, otherwise do not keep it. Step 12: Update the optimal route of the route population; Step 13: Have all routes been updated? If yes, jump to step 14; otherwise, jump to step 7; Step 14: Whether the termination temperature Tz is reached, if yes, the optimal route is output, otherwise, the cooling operation is performed and the process goes to step 6; Step 15: Based on the route obtained in step 14, use the A* algorithm to plan a path in the map.

Images